Research on overlapping community detection is still significant due to the rapid growth of different types of complex networks. Among the types of community detection, density-based clustering approaches did not get much attention in the field of overlapping community detection. We found two main drawbacks of the density-based approaches: 1) lack of research in computing the proper long-range distance (up to several hops) among the nodes in graph dataset; and 2) difficulty in computing overlapping belonging coefficients of the nodes in terms of diverse and irregular shaped clusters of graph datasets where distances among few nodes are available only (even not the positions of the nodes). In this article, we propose a novel overlapping community detection approach which we term as “density-based fuzzy community detection (DFCD)”, based on local densities of the nodes. Here, we capture the network topology in the form of bulk node-streams (node sampling) by employing biased random walk. We propose a novel way to compute long range distances among the nodes by utilizing the nodes-streams to find fixed number of global neighbors for each node, and accordingly we compute local densities of the nodes. We propose three relations between nodes and communities, where local densities and distances among the nodes are known only. Finally, we propose a novel belonging computation function based on the three relations, local densities, distances among the nodes. Our proposed approach is judged considering a number of real-life and synthetic datasets, where we consider ten state-of-the-art competitors. The experimental analysis shows superiority of the proposed approach over its competitors based on both accuracy and quality evaluations. Moreover, the detailed complexity analysis of the proposed approach shows linearly growth of the time complexity in terms of number of nodes and average degree, which is less than the same of majority overlapping community detection approaches.

Modern neural networks, despite their impressive accuracy, frequently suffer from miscalibration, yielding confidence estimates that poorly reflect true performance. Traditional approaches to calibration typically rely on post-hoc adjustments, which do not influence the intrinsic learning dynamics of the model. In contrast, this paper proposes a novel framework that integrates a self-deception - inspired mechanism into the training process. Drawing on insights from human cognitive processes - where controlled overconfidence and self-deception facilitate perseverance and exploration - we introduce an auxiliary module that selectively boosts the confidence of a network's predictions during training. This controlled confidence boost not only smooths the loss landscape by amplifying gradient signals in regions of ambiguity but also promotes adaptive exploration and robust optimization. Our theoretical analysis demonstrates that, under mild assumptions, the proposed mechanism can enhance learning dynamics and improve calibration without incurring significant computational overhead. By bridging cognitive theory with deep learning, our approach challenges the conventional view that overconfidence is inherently detrimental and paves the way for the development of more resilient and trustworthy AI systems.

Over the last few years, deep belief networks (DBNs) have been extensively utilized for efficient and reliable performance in several complex systems. One critical factor contributing to the enhanced learning of the DBN layers is the handling of network parameters, such as weights and biases. The efficient training of these parameters significantly influences the overall enhanced performance of the DBN. However, the initialization of these parameters is often random, and the data samples are normally corrupted by unwanted noise. This causes the uncertainty to arise among weights and biases of the DBNs, which ultimately hinders the performance of the network. To address this challenge, we propose a novel DBN model with weights and biases represented using fuzzy sets. The approach systematically handles inherent uncertainties in parameters resulting in a more robust and reliable training process. We show the working of the proposed algorithm considering four widely used benchmark datasets such as: MNSIT, n-MNIST (MNIST with additive white Gaussian noise (AWGN) and MNIST with motion blur) and CIFAR-10. The experimental results show superiority of the proposed approach as compared to classical DBN in terms of robustness and enhanced performance. Moreover, it has the capability to produce equivalent results with a smaller number of nodes in the hidden layer; thus, reducing the computational complexity of the network architecture. Additionally, we also study the sensitivity analysis for stability and consistency by considering different membership functions to model the uncertain weights and biases. Further, we establish the statistical significance of the obtained results by conducting both one-way and Kruskal-Wallis analyses of variance tests.

With ever-increasing abundance of text data, decision-making problems are becoming more complex. Such complexity is often a consequence of nuanced input linguistic information, which is already highly uncertain and subjective. In this article, a novel multiattribute decision-making (MADM) model named MARSHAL is introduced in order to capture the aforementioned nuanced characteristics from raw input linguistic data. It is for the first time in the literature of fuzzy based linguistic MADM models that MARSHAL treats inputs as rich linguistic features from a pretrained deep learning based large language model. This is in addition to the input being uncertain and hesitant, while also presenting vector arithmetic based information elicitation from corresponding hesitant fuzzy linguistic term sets. The idea is to introduce semantically-aware probabilistic attribute weights based on high-dimensional linguistic features learned by an enhanced variant of BERT, utilized to solve MADM problems alongside risk and regret-aversive behaviors of experts. The proposed model is tested for applicability on a real case-study of employee flight risk detection. Additionally, extensive quantitative and qualitative experiments are performed to prove the proposed model’s gained interpretability and adaptability among several other properties that existing congeneric models do not possess.

Diabetes mellitus (DM) is a disease impacting the regular activities and lifestyle of a majority of the human population throughout the world. Due to its adverse effects on health, regular interactions with healthcare providers become essential for early diagnosis, prognosis, and proper treatment plans. Such interactions often involve qualitative information, which when considered promotes effective and practical diagnosis. This is achieved in this article through the proposal of PerCIST, which is a perceptual computer-based nonclinical diagnostic tool for the prediction of DM. The proposed model is trained using cognitive understandings of individuals and takes into account the qualitative responses toward the parameters: age, body mass index, fasting blood sugar levels, HbA1c results, and heredity. These parameters, which are usually taken into account by medical practitioners for the diagnosis of DM, are decided through consultation of experts. The proposed model of PerCIST attains an accuracy of 93%, which is higher than any existing congeneric model, when tested on multiple individuals. Additional statistical and empirical experiments also highlight the prowess of the proposed model based on the significance of the chosen diagnostic parameters. Consequently, web and Android based applications, made freely available, are also provided to benefit people getting diagnosed with DM at home, especially during times of a pandemic when the entire world is under lockdown.

This paper proposes a novel fuzzy community detection (FCD) approach, which we term as ‘Label Propagation-Based Fuzzy Community (LaProFC)’, and shows that it has the ability to outperform the existing FCD approaches. While designing the proposed FCD approach, we introduce a new compound type similarity metric termed ‘proportion of common neighbors and edges-based similarity (CCS)’ to compute similarity between two neighboring nodes. By executing local exploration on graphs with modified local random walk (mLRW), most similar neighbors of each node are identified; and based on the directions of most similar neighbors some tentative communities are generated. Afterward, these tentative communities are corrected and stabilized by iteratively computing membership degrees of each node using a novel label propagation-based membership computation function. We also propose a novel edge-density-based technique called ‘community-weight based tie-breaking (CTB)’, which is incorporated with the membership degree computation function. We conduct extensive experiments with both real-life and synthetic datasets and show the working of the proposed approach. Our Proposed LaProFC approach outperforms baseline approaches in terms of popular quality and accuracy metrices including modularity and normalized mutual information. Further, popular multi-criteria decision making (MCDM) tools are used to show supremacy of the proposed approach by computing the ranks of different approaches through two sets of accuracy and quality metrices. Our proposed LaProFC approach supersedes other approaches in terms of faster computations and asymptotic time complexity.

This paper proposes a novel Deep Belief Network (DBN) architecture, the ‘Interval Type-2 Fuzzy DBN (IT2FDBN)’, which models the weights and biases with IT2 FSs. Thus, it introduces a novel algorithm for augmented deep leaning, which has the capability to address all the limitations of the classical DBN (CDBN) and T1 fuzzy DBN (T1FDBN). We comparatively evaluate the performance of the IT2FDBN by conducting experiments using the popular MNIST handwritten digit recognition datasets. Additionally, to demonstrate its robustness and generalization capabilities, we also conduct experiments taking two noisy variants of MNIST dataset, viz. the MNIST with AWGN (additive white Gaussian noise) and the MNIST with motion blur. We conduct extensive simulations by considering different combinations of nodes in the hidden layers of the DBN for better model selection. We thoroughly compare the results using well-known performance measures such as root mean square error (RMSE) and Error rate. We show that, in terms of RMSE values and error rates, the proposed IT2FDBN outperforms both T1FDBN and CDBN across all the three datasets. Further, we also provide the results of convergence, runtime-based comparison, and statistical analysis in support of our proposal.

Fuzzy community detection (FCD) aims to reveal the community structure by allocating quantitative values to nodes across different communities. This article proposes a fast FCD approach called the Expandable Local Community based Fuzzy Community (XLoCoFC) detection method based on max-membership degree propagation (max-MDP) and normalized peripheral similarity index ($ \boldsymbol{n}\mathbf{P}\mathbf{S}\mathbf{I}$). Initially, nodes having comparatively higher$ \boldsymbol{n}\mathbf{P}\mathbf{S}\mathbf{I}$values are considered as topologically dominating nodes and selected as seeds. For an initial community, called local community, seed’s$ \boldsymbol{n}\mathbf{P}\mathbf{S}\mathbf{I}$values from the respective neighbors’ peripheries are utilized as the neighbors’ membership degrees. Then an iterative process propagates max-membership degrees from nodes to nodes, and$ \boldsymbol{n}\mathbf{P}\mathbf{S}\mathbf{I}$values are used as factors in the propagation. In this propagation, local communities having more dominating nodes expand and others contract. The propagation process converges very quickly. Such simplicity in its design makes our proposed XLoCoFC approach to be very fast in finding community structures on large networks. Time complexity of the proposed approach is$ \boldsymbol{O}\left(\boldsymbol{n}\boldsymbol{d}^{2}\times \mathbf{lo}\mathbf{g}_{2} \boldsymbol{d}+\mathbf{k}\mathbf{l}\mathbf{q}\right)$which is significantly less than the majority of the FCD algorithms, for whom it is either$ \boldsymbol{O}\left(\boldsymbol{n}^{2}\right)$or more. Moreover, XLoCoFC has no dependence on any network feature. It does not require tuning of any parameter which may impact its output. To demonstrate the working of the proposed XLoCoFC approach, we conduct extensive performance analysis comparatively by executing a set of existing approaches on several popular real-life and synthetic networks with number of nodes ranging from 24 to 1134 890. Evaluation of the results considering the accuracy and quality metrics as well as a group MCDM technique clearly establishes the superiority of our approach over others.

Among the available processor-level energy savings schemes, dynamic voltage and frequency scaling (DVFS) is very popular and effective due to its widespread cross-platform use in designing energy-efficient scheduling algorithms. However, rapid frequency switching by DVFS based algorithms while minimizing the energy consumptions may result transient failures in the system. To avoid such failures and their catastrophic consequences, energy-efficient scheduling algorithms with the capabilities to provide more reliable task schedules are always in demand. Therefore, this paper introduces two novel low complexity energy-efficient task scheduling algorithms for heterogeneous computing environments. We term the first algorithm as ‘binary search-based energy-efficient scheduling with reliability goal (BSESRG)’ for running parallel task graphs in heterogeneous computing systems. We show that the proposed BSESRG has the capability to reduce energy consumption, and shorten the total schedule length by meeting the reliability goals upto a certain threshold. Then, we present our second algorithm, the ‘binary search-based energy-efficient fault-tolerant scheduling with reliability goal (BSESRG-FT), which ensures meeting the reliability goals with simultaneous consideration of fault tolerance. The proposed BSESRG-FT is able to reach higher reliability goals, reduce energy consumption, and shorten the total schedule length of a parallel task graph on heterogeneous platforms. We demonstrate the working of both BSESRG and BSESRG-FT through simulation experiments considering real-world task graphs, and show the supremacy of the two proposed algorithms over their respective peers (viz., ESRG and EFSRG) in terms of energy savings, schedule lengths, run times and reliability goals. The superiority of the proposed BSESRG and BSESRG-FT over their respective competitors are also validated on the real benchmark MiBench. Moreover, from the complexity analysis, we respectively find the time complexities of BSESRG and BSESRG-FT as$O\mathbf {(|\mathcal {X}|\times |P| \times log_{2}|F|)}$and$O\mathbf {(|\mathcal {X}|\times |P|^{2}\times log_{2}|F|)}$confirming their better computational efficiency than the respective peers.

In an Industry 4.0 ecosystem, all the essential components are digitally interconnected, and automation is integrated for higher productivity. However, it invites the risk of increasing cyber-attacks amid the current cyber explosion. The identification and monitoring of these malicious cyber-attacks and intrusions need efficient threat intelligence techniques or intrusion detection systems (IDSs). Reducing the false positive rate in detecting cyber threats is an important step for a safer and reliable environment in any industrial ecosystem. Available approaches for intrusion detection often suffer from high computational costs due to large number of feature instances. Therefore, this paper proposes a novel unsupervised IDS for Industry 4.0 which we term as: Unsupervised Intrusion Detection System for Industry 4.0 (UInDeSI4.0). We have substantiated the proposed UInDeSI4.0 approach through its experimentation on the well-known UNSW-NB15 Industry 4.0 dataset. The proposed UInDeSI4.0 employs feature selection approaches to obtain minimal and optimal features. These features are then used to train isolation forest to detect network traffic threats in an unsupervised manner. Accordingly, the proposed UInDeSI4.0 approach can efficiently differentiate between the normal events and the attacks or intrusions in environments with no label information. Experimental results show that the proposed UInDeSI4.0 provides better accuracy (∼63%) and a minimal feature set (nine) compared to traditional IDSs. In contrast to deep learning approaches, UInDeSI4.0 generates faster results with minimum features. In conclusion, we establish the superiority of UInDeSI4.0 approach as an accurate and computationally efficient IDS for Industry 4.0.

With the spread of the deadly coronavirus disease throughout the geographies of the globe, expertise from every field has been sought to fight the impact of the virus. The use of Artificial Intelligence (AI), especially, has been the center of attention due to its capability to produce trustworthy results in a reasonable time. As a result, AI centric based research on coronavirus (or COVID-19) has been receiving growing attention from different domains ranging from medicine, virology, and psychiatry etc. We present this comprehensive study that closely monitors the impact of the pandemic on global research activities related exclusively to AI. In this article, we produce highly informative insights pertaining to publications, such as the best articles, research areas, most productive and influential journals, authors, and institutions. Studies are made on top 50 most cited articles to identify the most influential AI subcategories. We also study the outcome of research from different geographic areas while identifying the research collaborations that have had an impact. This study also compares the outcome of research from the different countries around the globe and produces insights on the same.

The recently introduced ‘Improved Probabilistic Intuitionistic Fuzzy c-Means algorithm (IPIFCM)’ provides a Probabilistic Euclidean Distance Measure (PEDM) based computationally efficient clustering technique. Since IPIFCM is defined on Type-1 Atanassov Intuitionistic Fuzzy Sets (AIFS), it is unable to capture the uncertainty of the membership and non-membership values of a given datapoint induced by the hesitancy factor of the AIFS. Interval Type-2 Fuzzy Sets (IT2 FSs) deals with the uncertainty in the membership values. In this paper, we incorporate IT2 FSs in the IPIFCM algorithm by introducing upper bound and lower bound of the membership (and non-membership) values of each datapoint to model the change caused by the hesitancy factor. Accordingly, this paper proposes the ‘Probabilistic Interval Type-2 Intuitionistic clustering algorithm’ (P-IT2IFCM), which uses the interval probabilistic weights for PEDM to propose ‘Interval Type-2 Probabilistic Euclidean Distance Measure’ (IT2PEDM). The proposed algorithm provides superior results to existing Fuzzy c-Means (FCM) algorithms such as the basic FCM algorithm, IFCM algorithm, Kernelized-IFCM algorithm and IPIFCM algorithm when executed over various benchmark UCI datasets.

The recently reported ‘Improved Probabilistic Intuitionistic Fuzzy c-Means (IPIFCM) algorithm’ is a computationally efficient algorithm that does fuzzy clustering based on Probabilistic Euclidean Distance measure (PEDM). A significant issue with the IPIFCM algorithm is that it does not consider the hesitation factor's effect while updating the membership of a datapoint for a given cluster. Therefore, the convergence of the algorithm is not optimal. In this paper, we modify the membership function by adding the hesitation component to the IPIFCM clustering algorithm's objective function to propose 'Modified Improved Probabilistic Intuitionistic Fuzzy c-Means' (MPIFCM) clustering algorithm. The proposed MPIFCM algorithm helps in achieving a realistic clustering of the datapoints. This modification leads to the improvement in the accuracy as well as the convergence rate of the algorithm. Experiments over various benchmark UCI datasets confirm that our proposed algorithm provides better performance over its existing counterparts. Popular performance metrices such as accuracy, convergence rate, partition coefficients and cluster entropy are considered for comparative analysis of the performances of the studied algorithms.

The current ongoing pandemic COVID-19 situation severely impacts the tourism sector. The principal economic component in some of the countries is their tourism sector. So, a proper strategy to recommence the tourism sector needs to be formulated. We chalk out a plan of action while solving multiple criteria sorting (MCS) problem. The dealing of COVID-19 involves hesitancy and uncertainty thus, Atannasov’s intuitionistic fuzzy set is used to model this situation. The paper introduces an intuitionistic fuzzy grey relational analysis sort (IFGRA-sort) technique to strategize the reopening of the tourism industry. The proposed technique successfully solves the tourism industry problem given in Current Issues in Tourism (2021): 1-11, Taylor and Francis.

This paper reports a novel fuzzy community detection (FCD) algorithm, which we term as ‘Boosted NeSiFC (bNeSiFC)’, based on an improvement of the recently proposed NeSiFC approach. Similar to the basic NeSiFC approach, the proposed bNeSiFC also computes the similarity between two neighbors using the modified local random walk (mLRW). In the proposed bNeSiFC, a new similarity metric termed EDS is introduced to compute the pair-similarity for constructing the transition probability matrix of mLRW. The boosted NeSiFC outperforms over the basic NeSiFC in terms of a faster computation in finding the most similar neighbors through the incorporation of the newly proposed metric EDS. Also, we introduce a novel fuzzy membership degree computation method for the proposed bNeSiFC, which is much clearer and easy to interpret than the one used for the basic NeSiFC. Comparative analysis of the experimental results with eight different real-life datasets establishes the superiority of the bNeSiFC over the NeSiFC and other existing approaches.

This article proposes a neighbors’ similarity-based fuzzy community detection (FCD) method, which we call “NeSiFC.” In the proposed NeSiFC approach, we compute the similarity between two neighbors by introducing a modified local random walk (mLRW). Basically, in a network, a node and its’ neighbors with noticeable similarities among them construct a community. To measure this similarity, we introduce a new metric, called the peripheral similarity index (PSI). This PSI is used to construct the transition probability matrix for the mLRW. The mLRW is applied for each node until it meets a parameter called step coefficient. The mLRW gives better neighbors’ similarity for community detection. Finally, a fuzzy membership function is used iteratively to compute the membership degrees for all nodes with reference to existing communities. The proposed NeSiFC has no dependence on the network characteristics, and no adjustment or fine tuning of more than one parameter is needed. To show the efficacy of the proposed NeSiFC approach, we provide a thorough comparative performance analysis considering a set of well-known FCD algorithms viz., the genetic algorithm for fuzzy community detection, membership degree propagation, center-based fuzzy graph clustering, FMM/H2, and FuzAg on a set of popular benchmarks, as well as real-world datasets. For both disjoint and overlapping community structures, results of various accuracy and quality metrics indicate the outstanding performance of our proposed NeSiFC approach. The asymptotic complexity of the proposed NeSiFC is found as O(n2).

The Reliability Redundancy Allocation Problem (RRAP) is generally classified as a NP-hard problem. The RRAP aims to achieve the system's optimal reliability considering the minimum number of redundant system components while keeping volume, weight, and cost in mind as the key constraints. However, the dynamic changes in the manufacturing process lead to uncertainty in these system parameters, which makes them the suitable fit for being formulated as a fuzzy quantity. There are previous studies that have modelled the parameters as fuzzy and solved various fuzzy RRAP as separate problems to cope with the uncertainty. However, owing to some similarities between the cases, such as fuzzy series and complex bridge systems, they can also be solved simultaneously. As a result, using the Multi-factorial Evolutionary Algorithm (MFEA) method, this paper proposes a technique for simultaneously solving two fuzzy RRAP cases: fuzzy complex (bridge) and fuzzy series system. The similar characteristics of these two systems facilitate the evolution process through implicit knowledge. The experiments' findings demonstrate that the developed methodology, as compared to other evolutionary approaches using a benchmark dataset, was able to solve these problems effectively.

During the designing phase of real-time embedded systems (RTESs), available information on task characteristics is either incomplete or imprecise. So, task timing constraints are mostly approximated estimations by designers. This indicates that there are underlying uncertainties in these timing constraints, which can be appropriately modeled using fuzzy numbers. Moreover, feasible scheduling of tasks and energy efficiency are two essential requirements for better utilization and durability of RTESs. Thus, justifiable performance of this kind of systems warrants energy savings amid timely production of the computational outputs, although these two issues are mutually contradictory. This article reports a novel formulation of the energy efficient real-time scheduling problem in a fuzzy uncertain environment and proposes a novel solution approach called “ε-constraint coupled energy efficient genetic algorithm (ε-EEGA).” The working of the proposed approach is demonstrated taking a real-life example. Also, a thorough comparative analysis is provided considering well-known existing approaches including multiobjective evolutionary algorithms. Results are compared using popular performance metrics, which suggests that the proposed ε-EEGA is efficient in giving better energy savings with faster computations than its existing counterparts. Standard statistical tests such as analysis of variance and Kruskal-Wallis are performed.

Linguistic summarization is an approach of extraction of knowledge or linguistic patterns from datasets. Since summaries are brief, they promote quick analysis of data. Numerous researchers have utilized this approach to produce summaries which are easy to comprehend. Furthermore, linguistic summarization has been used to generate IF-THEN rules in the literature which not only convey data easily but is utilized in making decisions. In this paper we follow this approach to generate IF-THEN rules for diabetes on a dataset. This constructed dataset consists of responses from individuals for five parameters, crucial in the diagnosis of diabetes. Consequently, the quality of the rules produced using linguistic summarization is checked by the four quality measures namely: degree of truth, coverage, reliability and outliers. Among these, the degree of reliability is useful to find rules that represent dataset completely, and the outliers are used to find rules that deviate from the original result. Our experiment reveals results that are promising when compared to the PIMA dataset.

A perceptual computer generates appropriate word recommendations, rankings and/or classifications for any given application context. During the preparation of the codebook, perceptions of individuals are obtained in the form of end-points of intervals, followed by the selection of words from the codebook at a later stage. Keeping in mind the fact that humans are hesitant while providing responses, the perceptual computing paradigm does not handle such hesitancy at any stage. To conquer this deficiency within the framework of perceptual computing, we propose the idea of perceptual computing with comparative linguistic expressions. Through this proposal, we engage the hesitancy faced by people at different stages of the perceptual computing framework. We also introduce two different types of hesitancies that can occur, depending on the types of elements.

In this paper, Interval-valued intuitionistic fuzzy set (IVIFS) is exploited to propose a generalization of fuzzy TOPSIS method. We have termed the proposed TOPSIS method as Interval-valued intuitionistic fuzzy TOPSIS method (IVIFSTOPSIS). Here, the IVIFS-TOPSIS handles the Supplier Selection problem in which linguistic variables based criteria description is given. The best supplier obtained by IVIFS-TOPSIS is in concurrence with the other well-known fuzzy TOPSIS methods. It is possible to use IVIFS-TOPSIS over other types of linguistic variable multi-criteria decision making (MCDM) problems.

Recently proposed Probabilistic Intuitionistic Fuzzy c-Means Algorithm (PIFCM) is a Probabilistic Euclidian distance measure (PEDM) based clustering technique, which incorporate computation of probabilistic intervals (Pij, Qij) for each of the data point. PIFCM algorithm employs a random membership function $\frac{1}{{\left| x \right|}}$ and discards a data point if its membership value is uniformly distributed in the clusters. Fuzzy clustering always gets affected by the choice of the membership function. Accordingly, in PIFCM algorithm, membership function changes the properties of the data limiting its capabilities in giving consistent clustering results. Moreover, PIFCM algorithm incorporates computation of redundant matrices while finding Pijand Qij. In this paper, we propose some novel changes in the existing PIFCM algorithm, and hence introduce our Improved PIFCM algorithm. The improved PIFCM algorithm considers the min-max normalization as membership function, and also removes the redundant matrix computation that was used to find the Pijand Qijin the original PIFCM. Results over various UCI datasets validates the superiority of our improved PIFCM algorithm over FCM algorithm, IFCM algorithm and PIFCM algorithm.

Most of the time membership value in the fuzzy set cannot be exactly defined. Interval-valued fuzzy set (IVFS) is a special type of type-2 fuzzy sets which represents the membership value of the fuzzy set as an interval. IVFS assumes that membership interval can better represent the uncertainty in the data. Accordingly, IVFS can be used to obtain good clustering results since it can represent the uncertainty more appropriately. Thus, this paper proposes the interval-valued fuzzy c-means algorithm (IVFCM) which uses IVFSs to represent the data. The concept of the proposed IVFCM is then extended to introduce the interval-valued density based fuzzy c-means (IVDFCM) algorithm based on the distance measure of IVFSs. Both IVFCM and IVDFCM are simulated over various UCI benchmark datasets to show their suitability and supremacy over their existing counterparts.

The rapid growth in industrialization has proportional effect on the increase in carbon emission as well as economic growth of a nation. Nevertheless, there are many nations with unavailable information on their gross domestic products (GDPs). Therefore, primarily, this paper addresses the problem of predicting missing GDP of these nations with the help of their carbon emission data. However, the available data of these countries are insufficient for training a predictive machine learning model. So, we have focused on the emerging yet under-explored area of multi-source unsupervised transfer learning to enlarge the training domain by introducing the detection and removal of anomalies in order to build a robust prediction framework. This is empirically evaluated over the carbon emission and per capita GDP data, collected from the World Bank repository, of a number of developing countries as well as over a set of mixed (developed and developing) countries. Five different domains generated using multi-source unsupervised transfer learning framework are evaluated using three different machine learning models. The best among them is then used to predict the missing per capita GDP of a nation.

Present-day handheld battery-enabled devices such as smartphones and tablets attract rich user experience but are often criticized for their short battery lives. Battery life is a subjective term and depends on a user's perceptions. A novel work to achieve power optimization for these devices, according to users' perceptions, was the design of user-satisfaction-aware power management approach, perceptual computer power management approach (Per-C PMA). But we have found that the design of Per-C PMA requires collection of data intervals from a group of subjects. This limits the practical viability of Per-C PMA for highly personal handheld battery-enabled devices such as smartphones and tablets. So, here we propose a user-satisfaction-aware PMA called Per-C for Personalized Power Management Approach or “Per-C PPMA,” one that achieves significant reductions in power consumption compared to existing PMAs and noticeable improvements in the overall user satisfaction. Per-C PPMA uses the mathematical technique of person footprint of uncertainty (FOU) to process users' linguistic opinions. Person FOU can either use an interval approach (IA) or Hao-Mendel approach (HMA) for data processing. The recommendations generated using IA and HMA are the same. However, IA takes a much higher computational time than HMA, even though both have the same asymptotic complexity of O(w*n). We strongly believe that Per-C PPMA is a novel technique and our work is the first such application of Person FOU on any hardware platform. An important outcome of this study is a ready-to-use mobile app “Per-C PPMA” (currently freely available on the website http://www.sau.int/~cilab/).

This paper deals with one of the well-studied research problem known as reliability redundancy allocation problem (RRAP). RRAP is finds an optimal balance in selecting reliability of the components and the number of redundant components considering the effect of cost, weight, and volume on the overall system as constraints. Reliability is a very crucial issue in every engineering application. In this paper, our study aims to find the optimal layout for RRAP using brain storm optimization algorithm in objective space (BSO-OS). BSO is inspired by human cognitive skills. The convergent and divergent operation to reduce the search space and diversification of population are the prime objective in BSO. Results are compared by solving the RRAP with the self-organizing migrating algorithm (SOMA) and the genetic algorithm (GA). It is found that BSO-OS outperform others in terms of time complexity providing similar results.

Increase in transportation has led to alarming pollution globally causing an adverse effect on the environment. Pollution-Routing Problem (PRP), a version of well-known Vehicle Routing Problem, has attracted researchers to develop efficient solutions to cut-off fuel consumption and minimize greenhouse gas emissions. This paper presents a novel effort with an objective pertaining to the minimization of fuel consumption (CO2emissions). Thus, we have modified a recently developed Brain Storm Optimization in Objective Space (BSO-OS) algorithm to design a solution technique for PRP. Further, we have incorporated a recombination operator and a method to produce diverse solutions into the existing BSO-OS. The objective defined here is based on the load factor; thus, our modified BSO-OS concentrates towards deducting the load and minimize the fuel consumption efficiently. For comparison, we have also implemented a Genetic Algorithm (GA) for PRP. The comparative study of the experimental results demonstrates the efficacy of the algorithms for the addressed problem. BSO-OS outperforms GA in terms of running time and generates a varied range of solutions.

A number of real-life scenarios involving decision making may be modelled as optimization problems. In these optimization problems, the human preferences and thinking constrain achieving the optimal value of the problem objective(s). If there is a single objective, then the optimization problems are called single objective optimization problems (SOOPs) else the multi-objective optimization problems (MOOPs). Various solution methodologies have been proposed for SOOPs and MOOPs, which are useful, as long as the SOOPs and MOOPs involve the numeric data. However, the data is generally in linguistic form (or words), when elicited by the human beings. Therefore, the SOOPs and MOOPs are referred as single objective linguistic optimization problems (SOLOPs) and multiobjective linguistic optimization problems (MOLOPs), respectively, in such situations, to emphasize the existence of linguistic information in optimization problems. In these LOPs, the value of the objective function(s) may not be known at all points of the decision space, and therefore, the objective function(s) as well as problem constraints are linked through if- then rules. Previously, the Tsukamoto's inference method was used to solve these types of LOPs; however, it suffers from drawbacks. As, the use of linguistic information inevitably calls for the utilization of computing with words (CWW), hence, in this paper, we discuss the solution methodologies for LOPs based on the perceptual reasoning (PR). PR is a novel CWW engine design for the CWW approach of perceptual computing. We also demonstrate the applicability of PR based solution methodology for LOPs to the case study of car purchase modelled as LOP.

Linguistic information is encountered frequently in the real-life situations, especially those involving human beings. The linguistic information is vague and imprecise; however, human beings compute seamlessly using it. Use of linguistic information inevitably requires the use of computing with words (CWW) methodology for its processing, in a manner similar to a computer. In a number of recent works, it was shown that the CWW approach of perceptual computing (Per-C) is better at processing the linguistic information and generation of unique recommendations, in comparison to other CWW approaches. Furthermore, in another work, a formulation of Per-C was used to propose a novel solution methodology for the multi-objective linguistic optimization problems (MOLOPs), where a novel design of Per-C's CWW engine was used called the perceptual reasoning (PR). In this PR based solution methodology for MOLOPs, the codebook was generated using the process of data collection. However, we fell that there can be scenarios, where it is not possible to collect the data for construction of the codebook. Therefore, in this paper, we propose to use the linguistic terms with symmetric membership functions and distributed uniformly on the information representation scale for such scenarios. Furthermore, the uniqueness of this work is that only the middle term's location from the linguistic term set needs to be established initially. Other linguistic terms are generated using the various operations of the linguistic hedges. We have also demonstrated the applicability of the PR based solution methodology for MOLOPs (based on symmetric term sets), to the case study of pollution routing problem, modeled as a MOLOP.

Deep learning (DL) has played a crucial role in many domains of image and pattern recognition, extraction of features from video and text processing etc. One of the quintessential elements of deep learning is Restricted Boltzmann Machines (RBM). RBMs are capable of extracting the high-level features from raw data very efficiently. Nevertheless, feature extraction process is prone to external and unwanted noises, which introduces uncertainty in the decision making process. Moreover, existing RBM-based DL methods are not robust enough to handle such noises in the data samples while training within layers. To tackle these drawbacks, Fuzzy Restricted Boltzmann Machine (FRBM) had been available in the literature. FRBM utilizes Type-1 Fuzzy Sets (T1FS) to handle such uncertainties in governing parameters of the system. However, membership values of membership functions used in T1FSs are also crisp. Thus, in this state-of-art paper, we propose the use Interval Type-2 Fuzzy Sets (IT2FSs) to model parameters in RBM for training, as they are efficient in handling higher level of uncertainty. Experiments performed for MNIST digits show more generative and discriminative capabilities of IT2FRBM over RBM and FRBM.

This paper proposes the novel way to estimate the gross domestic product (GDP) of a country from its carbon emission (CO2) data. This alternative method to predict GDP is required for the war affected and non-accessible nations as the macroeconomic data available for those nations is highly unpredictable or insufficient. However, first we need to train and develop a reliable model for which we have used the Transfer learning (TL) approach. TL is applied in a way that a neural network (NN) or machine learning (ML) model is trained on the developed nation's data and used for testing with developing nation data. The NN models used in this paper are Extreme Learning Machine (ELM) and Generalized Regression Neural Network (GRNN), while ML model is Support Vector Regression (SVR). Since the data is measurement data collected from several devices, it contains noise and hence, uncertain. Thus, first we have modelled the dataset with type-1 fuzzy sets and then with the interval type-2 fuzzy sets. The results are then compared with the crisp input data values.

Owing to the increasing complications in our daily lives, psychological problems too are on the rise. But to be able to lead a healthy life is the foremost priority for any human. To do so, proper diagnosis must be done. Since very long time, psychometric tests serve the purpose of providing personalized feedback about the possible existence of clinical symptoms for any kind of psychological illness. In this era of advancing computing technology, many computer based psychometric tests have been presented, which have numerous advantages. However, they are mostly based on the typical tick-the-appropriate-option pattern. Given the fact that humans find it inherently easier to respond with linguistic information, we feel that modelling the response of the test takers might result in better understanding of the their conditions. Also, there might be different forms of questions requiring responses from non-homogeneous domains of information. One must also keep in mind the hesitancy that humans face while zeroing in on a single linguistic term for their response. Shedding light on all the aforementioned points, we propose in this paper, a new linguistic decision making model based on non-homogeneous hesitant fuzzy information (LDM-NHFI) for psychometric tests. We demonstrate the suitability of the proposed model on a sample psychometric test towards the end of the paper. The presented model is novel because to the best of our knowledge, no such approach has been made to handle linguistic responses in psychometric tests.

In this paper, we have proposed an archived simulated annealing based novel approach for solving multi-objective energy-efficient scheduling on heterogeneous DVS activated processors in high-performance real-time systems. Real-time task scheduling problem is a well-known NP-hard problem. In these systems, tasks are usually associated with deadlines and represented by directed acyclic graphs since they depend on each other. So, system designers face difficulty in finding suitable solutions that can satisfy all the objectives of task scheduling, as warranted for proficient operations of such systems. Hence, this paper introduces a novel algorithm, called archived multi-objective simulated annealing for energy-efficient real-time scheduling (AMOSA-E2RTS) that finds an optimal schedule satisfying the precedence and deadline constraints. In the proposed algorithm, a domination concept leads towards finding the optimal trade-off solutions and tasks are prioritized according to three different policies i.e., latest deadline first (LDF), execution ranking and energy ranking policy. A suitable numerical example is used to demonstrate the working of the proposed approach. Experimental findings suggest that the proposed algorithm is capable of producing energy efficient scheduling decisions which satisfy all related constraints. Statistical analysis of the results has been conducted.

Image steganography is the art of hiding secret data into an image in such a way that it cannot be detected by any intruder. For steganography, image is a good carrier because it contains a higher redundancy of pixel values, making it less sensitive to the human visual system. Therefore, embedding in those pixels of the image leads to high visual quality, payload capacity and security. This paper proposes an interval type-2 fuzzy logic-based system to detect those pixels of the image which are less sensitive to human eyes. The embedding is performed on the selected pixels using the least significant bit (LSB) method. We have termed this method the interval type-2 fuzzy logic system based LSB (IT2FLS-LSB) steganographic method. The experimental simulations were performed on a collection of image datasets to show the efficacy of IT2FLS-LSB embedding. Quality index metrics such as PSNR (peak signal-to-noise ratio), UQI (universal quality index), and SSIM (structural similarity measure) are used to assure the visual quality of the stego images. We have also evaluated the high payload capacity of our proposed method.

Real life application domains frequently involve elicitation of information in the linguistic form, since human beings understand and express themselves using `words'. One such application domain is the decision making, on the basis of assessments about various alternatives by an evaluator. While providing the assessments about the alternatives, the evaluator is inclined to prefer one alternative over the other, to a certain degree. If the evaluator articulates the assessments in the linguistic form, then these assessments may be termed as linguistic preferences. One area where the linguistic preferences can find potential use is in the students' performances evaluation in a test or examination. In various examinations, a student's performance is assessed in comparison to that of others, based on his/ her score. A student with a comparatively higher score is a better performer than the one whose score is lower or vice versa, when compared quantitatively. However, the judgment about how better the performance of one student is, than the other, depends on the evaluator's perceptions. Thus, the faculty member or the evaluator expresses the student's performance using linguistic preference relations. In the present work, we propose a solution approach for assessing the students' performances based on evaluator's linguistic preferences using the consistency driven approach. The approach expresses the decision problems in the form of linear programming model and generates a ranking order of the alternatives. In the present problem, the consistency driven approach produces a ranking order of the students based on their respective performances.

Scheduling is the art of ordering a number of tasks with a fair allocation of resources with the intent of ensuring that these tasks are executed in minimal time. The processors, on which these tasks are executed, can be identical or nonidentical. They may have same or different execution capabilities. In real life situations, these tasks are in need of different types of resources such as network bandwidth, memory and processor speed. These resources are generally attached to the processors and the tasks are executed or assigned to the appropriate processor based on a scheduling algorithm. The scheduling algorithm determines the total execution time required to complete all the tasks all the processors and tries to minimize this total completion time. However, in real life, the values of these scheduling resources are uncertain and therefore they affect the total completion times to varying degrees. As human beings naturally understand and express themselves using words, therefore to incorporate human factors in these scheduling problems, the values of these criteria were specified linguistically in a recent work. The system was designed as a combination of if-then rules and Mamdani inference mechanism was used to determine the crisp value of completion time. Furthermore, the number of if-then rules required to design the system were very large. However, the solution to a problem formulated by the linguistic variables should be in linguistic form. Therefore, in the present work we propose the use of 2-tuple fuzzy linguistic approach to find the solution of parallel processor scheduling involving linguistic data values. The proposed approach is capable of giving linguistic solution using a small sized if-then rule base.

We often have many datasets where hard clustering algorithms do not deliver satisfactory clustering results. It is found that many times fuzzy clustering technique improves the clustering results obtained by hard clustering algorithms. Fuzzy c-means (FCM) is the most prominent fuzzy clustering techniques whose improvement was proposed through the introduction of intuitionistic fuzzy set (IFS) based c-means algorithm. In order to implement IFS based c-means algorithm over a real valued dataset, data points were first converted into IFSs by employing a highly popular technique known as Yager's generating function. The Yager's generating function tunes only the non-membership and hesitancy component of an IFS. Therefore, IFS based c-means algorithm produces compromised clustering results. In this paper, we have generalized the Yager's generating function in such a manner that our IFS generation function tunes all the three components of the IFSs. We have utilized the proposed IFS generation function in two highly used IFS based c-means algorithms of clustering known as intuitionistic fuzzy c-means (IFCM) and Novel intuitionistic fuzzy c-means (Novel-IFCM) algorithms on the UCI datasets. Our results obtained using the proposed function are better than the results obtained using YGF.

Supplier performance evaluation (SPE) is the task of evaluating the performance of a supplier based on various criteria for a particular application. Due to this, SPE is predominantly a multi-criteria decision making problem (MCDM). Since the problem of SPE is qualitative in nature, the decision makers evaluating the supplier find it convenient to express the assessments in the form of linguistic expressions. But there are situations where the decision makers are hesitant about expressing their opinions for the supplier which would enable them to provide their opinions through single linguistic term, for each supplier criterion. Therefore, the assessments are often specified as complex linguistic terms not present in the linguistic term set. To overcome the difficulty of eliciting linguistic information from these complex expressions, we propose to use hesitant fuzzy linguistic term sets (HFLTS) in the SPE problem. We show that the use of HFLTS provides advantages that are not possible with the linguistic 2-tuple model. Our work is novel because to the best of our knowledge, HFLTS have not been used for SPE before.

Decision making problems which are inherently hierarchical come under the category of multi-level programming problems. Whenever the hierarchy of problems is limited to two levels, the multi-level programming problem becomes a bilevel programming problem. Classical bilevel problems contain a leader problem subject to its constraints, and a follower problem subject to its own constraints. The follower problem acts as a constraint for the leader problem. However, in real practical cases, there may be a number of followers that constrain the solution space of the leader problem. Moreover, there may be certain impreciseness prevailing in the quantification of the resources of both the leader and the follower problems. In this paper, we deal with linear bilevel programming problems with multiple followers, which suffer from uncertainty, calling them fuzzy linear bilevel programming problems with multiple followers. While many solution approaches for such problems have already been introduced, we propose a novel defuzzification and K-th best algorithm based solution approach in this paper. Relevant results with a suitable numerical example have also been shown.

The CO2emission due to industrialization is a crucial parameter which is directly proportional to the economic growth of any country/nation. However, for the non-accessible nations and war-torn nations with highly unreliable or insufficient macroeconomic data, the prediction of gross domestic product (GDP) is a challenging task. Thus, in this paper, we have proposed a novel approach for the reliable GDP estimation utilizing only the CO2emission data. For this purpose, transfer learning (TL) is applied which learns on the previously acquired information and solve the new task. The training is performed on the GDP data of the developed nations and then prediction is estimated for the developing nations. This is implemented using kernel extreme learning machine (KELM) in which the output weights are modelled using interval type-2 fuzzy sets (IT2 Fss) for the effective transferal of knowledge from developed to developing nation. Experimental results have shown that the proposed IT2F-KELM provide much-improved RMSE as compared with the traditional KELM.

This paper presents a novel probabilistic similarity measure (PSM) for Atanassov intuitionistic fuzzy sets. It then exploits PSM to propose an adaptive probabilistic similarity degree and develops the novel probabilistic λ-cutting algorithm for clustering. Further, the probabilistic distance measure (obtained from the PSM) is used to develop a new clustering technique, which we have named “probabilistic intuitionistic fuzzy c-mean (PIFCM) algorithm”. Simulation experiments have been conducted over a variety of datasets including UCI machine learning datasets and realworld car dataset. The results obtained have been thoroughly compared with other well-known clustering techniques such as fuzzy c-mean (FCM), intuitionistic fuzzy c-mean, association coefficient method, and λ-cutting method. Based upon the experimental results, it can be concluded that our probabilistic λ-cutting algorithm and PIFCM algorithm outperform their existing counterparts.

The multiobjective reliability redundancy allocation problem (MORRAP) aims to ensure high system reliability in the presence of optimally redundant components. This is one of the most important design considerations for system designers. Due to the associated uncertainty in component parameters, precise computations of overall system reliability, cost, and weight, etc., are difficult during design time. Hence, these parameters are befitting to be modeled as fuzzy quantities. As type-1 fuzzy numbers have limitations in representing higher order uncertainties, so this paper models the component parameters viz., reliability, cost, and weight with interval type-2 fuzzy numbers. Thus, we propose a novel formulation of MORRAP, termed as interval type-2 fuzzy multiobjective optimization problem (IT2FMORRAP). A popular multiobjective evolutionary algorithm, viz., nondominated sorting genetic algorithm II, is used to solve the proposed IT2FMORRAP, for which we have developed two novel algorithms in this paper. Numerical examples are included to demonstrate the solution approach. On comparing the outcomes with earlier results, we have found that the proposed IT2FMORRAP outperforms classical as well as other type-1 fuzzy-number-based approaches.

Present day portable devices such as laptops, smartphones, etc., offer their users fastest processors, advanced operating systems, and numerous applications. However, a large section of the users are critical to the available battery capacity and its lifetime. This is because performance of the battery and its lifetime as perceived by the users are quite subjective in nature. It depends directly on user satisfactions, which are usually expressed in terms of words. So, in this paper, we propose a user-satisfaction-aware energy management approach, called “perceptual computer power management approach (Per-C PMA),” based on the technique of perceptual computing. At the heart of our technique is the perceptual computer that processes the linguistic input of the users to aid in the selection of a suitable processor frequency, which plays a significant role in the overall energy consumption of the systems. The Per-C PMA minimizes the energy consumption, while still keeping the user satisfied with the perceived system performance. The Per-C PMA achieves (1) reductions of 42.26% and 10.84% in power consumption, and (2) improvements in the overall satisfaction ratings of 16% and 10%, when compared to other existing power-saving schemes such as ON-DEMAND and human and application-driven frequency scaling for processor power efficiency, respectively. Per-C PMA is the first such application of Per-C on any hardware platform. It is implemented as Ubuntu scripts for end users and can be downloaded from: http://sau.ac.in/~cilab/. We have also provided the MATLAB files so that interested researchers can use it in their research. For the ease of the users, the Ubuntu scripts and the MATLAB codes are given in the graphical user interface mode; a demo video on how to use the software is also provided on the webpage.

In the modern era of industrialization and globalization, distribution and control of goods are essential aspects for multinational corporations and strategic partners. Vendor managed inventory (VMI) is one of the well-known strategies of merchandizing between supplier and retailer. In this paper, we consider different number of suppliers and retailers to perform business under VMI system and formulate three: single-supplier and single-retailer, single-supplier and multi-retailer, and multi-supplier and multi-retailer VMI systems. The objective is to minimize the total cost of VMI system. Since it is a non-linear integer programming problem, this paper proposes a novel hybrid biogeography-based optimization algorithm to solve it. We enhance the proposed algorithm by embedding stochastic fractal search (SFS) in biogeography-based optimization (BBO). SFS algorithm is a newly developed powerful evolutionary algorithm to find global optimum much faster and efficiently. The diffusion process of SFS improved the exploitation ability of search in BBO. Our proposed algorithm is applied on all three versions of VMI systems under different constraints. We have considered suitable input data for all the different problems and obtained the results. By comparison, we show that the results outperformed for all VMI systems.

Reliability redundancy allocation problem (RRAP) is an optimization problem with objective to maximize the system reliability considering component reliability and redundancies as decision variables. RRAP was mostly solved as a single level optimization problem. However, the nature of the problem fits quite well in the framework of bilevel optimization. In this paper, we have proposed two novel bilevel formulations for the RRAP and solve them using a latest bilevel optimization algorithm called BLEAQ (bilevel evolutionary algorithm based on quadratic approximations). So far we knew no other research has been reported till date, where RRAP was addressed with bilevel optimization algorithm. Here, optimization is needed at two separate levels, where one problem is encircled within another problem. The inner problem is known as lower-level problem and the external problem is called upper-level problem. Here, we have presented two mixed-integer non-linear bilevel formulations for the RRAP of series-parallel system in a competitive environment. The purpose of the upper-level problem is to determine the component reliability that maximizes the total system reliability; whereas, lower-level problem minimizes the total cost (or weight) needed. We demonstrate the applicability of our approach with a suitable numerical example and show that our proposed approach works quite well than existing single level optimization tools.

Vendor managed inventory (VMI) is one of the well-known strategies of merchandizing between suppliers and retailers. The essential aspect in VMI model is to fulfill the demand of the retailers by the suppliers. Since, the costs of items are highly volatile, it is not possible to design a VMI model with deterministic demand. In this paper, we have considered one supplier and retailer to perform business of multi-product under VMI model with uncertain demand. Thus, it proposes a novel VMI model, termed as `interval type-2 fuzzy vendor managed inventory (IT2 FVMI) model' that considers interval type-2 fuzzy number to represent the uncertain demand. The objective is to minimize the total cost of VMI by finding the optimum ordered quantity and backordered level. We develop a particle swarm optimization algorithm based solution approach to solve it. A suitable real-world application is considered to perform the simulation. We have considered different number of instances of the products to obtain the results. By comparing the results with the deterministic demand and type-1 fuzzy number, we show the efficacy of our proposed IT2 FVMI model.

Multi objective linguistic optimization is a useful mathematical technique to solve problems that interdependent criteria. In such problems, values of the objective functions may be unknown at some points, when the link between the variables and the objective functions are defined linguistically through if-then rules. While solving this type of problems, Tsukamoto based reasoning method has proved useful for converting objective function to a crisp form, and then using the resulting objective function to solve by any traditional optimization technique. However, this method suffers from a drawback that the resulting solution is in numeric form whereas it should have been in linguistic form, owing to the linguistic definition of if-then rules. So, here we propose 2-tuple fuzzy linguistic representation model based method for solving the Multi objective linguistic optimization problem. We demonstrate the novelty of our approach through a suitable example. We also prove that the proposed approach generates unique recommendation in linguistic form.

Pollution routing problem (PRP) is an NP-hard multi-objective optimization problem. The main goal is pollution reduction and secondary goals are cost/distance minimization, profit maximization etc. We have considered two unique models with two different set of objectives viz. (i) distance and fuel consumption, and (ii) weighted load and fuel consumption. Here, system parameters like demand, driver wages, timing constraints etc. can't be predicted a-priori and involve multiple opinions from the designers. Thus, such uncertain system parameters can be modelled using fuzzy sets. As type-1 fuzzy sets (T1 FSs) has limitations in modelling higher order uncertainty, this paper models these uncertain parameters with interval type-2 fuzzy sets (IT2 FSs). We have solved the problem by an efficient multi-objective evolutionary algorithm viz. NSGA-II (non-dominated sorting genetic algorithm-II). Numerical examples demonstrate the efficiency of the proposed technique over existing (crisp and type-1 fuzzy set based) approaches.

Restricted Boltzmann Machine (RBM) is a generative, stochastic neural network with two separate layers of hidden and visible units. Training data samples in RBM are usually corrupted by noise. RBM is not robust enough to handle such noises, which leads to uncertainty. In the literature, Fuzzy RBM (FRBM) has already been proposed for enhancing deep learning. In FRBM, the parameters of RBM are modelled as type-1 fuzzy numbers. However, there can be multiple sources of uncertainties such as noises in the data measurements, variations in the environment where they are deployed, etc. Such uncertainties cannot be modelled by type-1 fuzzy sets (T1 FS), since their membership values are themselves crisp in nature. On the other side, IT2 FS can model higher order uncertainties with their fuzzy membership grades. So, we propose to use interval type-2 fuzzy sets (IT2 FS) to model uncertain parameters of RBMs in the learning stage. Since deep neural network (DNN) is pre-trained using stacked RBMs, modeling noises using IT2 FS would demonstrate high performance and low root mean square error (RMSE) while learning. Thus, we propose a new algorithm viz. Interval type-2 fuzzy set based approach for enhanced deep learning (IT2 FS-EDL) in which RBM parameters are modeled as type-2 fuzzy sets in the learning process. Numerical examples and experimentations have been demonstrated to present the suitability of our proposed approach.

Optimal task allocation among the suitably formed robot groups is one of the key issues to be investigated for the smooth operations of multi-robot systems. Considering the complete execution of available tasks, the problem of assigning available resources (robot features) to the tasks is computationally complex, which may further increase if the number of tasks increases. Popularly this problem is known as multi-robot coalition formation (MRCF) problem. Genetic algorithms (GAs) have been found to be quite efficient in solving such complex computational problems. There are several GA-based approaches to solve MRCF problems but none of them have considered the dynamic GA variants. This paper considers immigrants-based GAs viz. random immigrants genetic algorithm (RIGA) and elitism based immigrants genetic algorithm (EIGA) for optimal task allocation in MRCF problem. Further, it reports a novel use of these algorithms making them adaptive with certain modifications in their traditional attributes by adaptively choosing the parameters of genetic operators and terms them as adaptive RIGA (aRIGA) and adaptive EIGA (aEIGA). Extensive simulation experiments are conducted for a comparative performance evaluation with respect to standard genetic algorithm (SGA) using three popular performance metrics. A statistical analysis with the analysis of variance has also been performed. It is demonstrated that RIGA and EIGA produce better solutions than SGA for both fixed and adaptive genetic operators. Among them, EIGA and aEIGA outperform RIGA and aRIGA, respectively.

In this paper, we have proposed a Bayesian optimization based novel approach for multi-objective task scheduling in real-time heterogeneous multiprocessor systems. Task scheduling problem in multi-processor real-time systems is a NP-hard problem. In such systems, scheduling of tasks becomes a huge challenge for the scheduler designers; especially when tasks are inter-dependent and have deadline constraints. Interdependent or precedence-constrained tasks are often represented as directed acyclic graphs. Most of the real life applications require state of the art planning and scheduling schemes for safer and efficient operations. Thus, we propose the algorithm `moBOA-RTS' (multi-objective Bayesian optimization algorithm for real time scheduling) to find an optimal schedule satisfying all the constraints within reasonable time. Here, learning of task graph is made through Bayesian networks. At first, tasks are allocated to different processors and then LDF (latest deadline first) based priority is used to determine the task execution on individual processors. The proposed approach can be applied on many processor real-time systems, where both the scenarios viz. homogenous and heterogeneous processing environments are prevalent. Experimental analysis shows that our approach produces optimal decisions for feasible scheduling that ensures the compliance of all real time and precedence related constraints.

Transfer learning framework is designed to use previously acquired knowledge to solve a new but somewhat related task (like humans do). Non-availability of sufficient and relevant information in building a learning model is a major bottleneck in this research area. However, such models are highly susceptible to negative transfer learning (NTL) during transferral of knowledge due to the hesitancy in the decision making. Negative transfer learning may cause chaotic learning and have a profound effect on their predictive precision. In this paper, we have proposed a novel Intuitionistic Fuzzy Domain Adaptation (IFDA) algorithm, which uses Yager-generating function over Atanassov's Intuitionistic fuzzy set theory in conjunction with modified Hausdorff Intuitionistic similarity metric to build a fuzzy domain adaptation algorithm which is independent of supervised machine learning technique. It exploits the hesitancy margin in intuitionistically fuzzified features by eradicating similar looking but useless instances. Therefore, it selects optimal source instances from a previous problem in bridging the knowledge gap, in order to solve a new target problem, by containing negative transfer learning.

Green Computing is the area that has proliferated rapidly in recent years. It has proved to be quite useful in reducing the power consumption related expenditures across all the sectors that are having information technology (IT) related services. Developments in the fields of VLSI and communications have provided us with miniature sized highly efficient computing platforms such as tablets, smart phones, etc. However, these devices are constantly criticized for their low battery lives. The concepts of green computing has the potential to extend the battery lives of these small sized computing devices. As more and more applications are being hosted by these modern day devices, our major aim is still to keep a user satisfied with the perceived system performance and elongate its battery life. Humans express in "words" and therefore we make use of perceptual computing to process user feedback about the perceived system performance. Smartphone or tablet is a personal device. Therefore we present in this paper an algorithm called 'TFA (Tablet Frequency Adviser)' that utilizes the concept of single person FOU. Here, the perceived system performance is measured by a single user and not a group of users. The proposed TFA algorithm ensures elongation of the battery life of a tablet with optimum user satisfaction. We have included the experimental results to justify the real-life applicability of the proposed TFA algorithm.

Multi-robot coalition formation (MRCF) problem deals with the formation of subsets of robotic to handle a particular task. In such a system, every task is executed by multiple robots. Thus, cooperation and coordination among the robots is very important. One of the key issues to be investigated for smooth operation of a multi-robot systems is finding an optimal task allocation among the suitably formed robot groups (sub sets). Considering the complete execution of available tasks, the problem of assigning available resources (robot features) to the tasks is computationally complex, which may further increase as number of tasks increases. Genetic algorithms (GA) have been found quite efficient in solving such complex computational problems. There are several algorithms based on GA to solve MRCF problems but none of them have considered the dynamic variants. Thus we apply immigrants based GAs viz. RIGA (random immigrants genetic algorithm) and EIGA (elitism based immigrants genetic algorithm) to optimal task allocation in MRCF problem. Comparative performance evaluation has been made with respect to SGA (standard genetic algorithm). Finally, we report a novel use of these algorithms making them adaptive with certain modification in their traditional attributes by adaptively choosing the parameters of genetic operators. We name them as aRIGA (adaptive RIGA) and aEIGA (adaptive EIGA). Simulations experiments have demonstrated that RIGA and EIGA produces better solutions then SGA in both the cases (with fixed and adaptive genetic operators). Among them, EIGA and aEIGA outperforms RIGA and aRIGA respectively.

In this paper, we have addressed the reliability-redundancy allocation problem with a particle swam optimization based technique. The parameters of the system components are actually imprecise or uncertain quantity since those are generally guessed by the designers during the design-time. Thus, important features of the designed system, viz. reliability, costs, weight etc very suitably qualifies to be considered as fuzzy quantity. Our problem formulation considers these parameters as type-2 fuzzy quantity. There are few reports where the problem has been studied under type-1 fuzzy uncertainty. As far as we know, no research has been reported where the problem has been addressed with a particle swam optimization based approach in a type-2 fuzzy environment. Suitable examples are included to demonstrate our approach. Results are compared showing that the type-2 fuzzy uncertainty based approach outperforms other recently reported results.

In this paper, we have reported a new approach for employing the non dominated sorting genetic algorithm-II (NSGA-II) with the type-2 fuzzy sets in optimizing energy in real-time embedded systems. The multi-objective problem of energy efficiency and timeliness of tasks has been extensively studied. Little variations in the task timing parameters produce considerable variations in the results of the critical real-time computations. Importantly, at the system designing phase these timing parameters are completely unquantifiable. We therefore propose here a new algorithm for real-time scheduling in type-2 fuzzy uncertain domains. We have included comparative results obtained from models with crisp timing parameters and their fuzzy type-1 and type-2 counterparts. From the observations of the outcome, it is found that model with crisp timing parameters gives the worst result as energy consumption in the system is maximum at a constant earliness. The crisp model is outperformed by both fuzzy type-1 and type-2 models and ensures significant reductions in energy consumption. Whereas fuzzy type-2 model overwhelms both fuzzy type-1 and crisp model in ensuring task completions with maximum earliness. Suitable numerical examples are included to demonstrate our proposed approach.

In Intuitionistic fuzzy sets(IFSs), experts assign both membership value and non-membership value to each fuzzy element x with a certain degree of hesitation. The hesitancy in the opinion of the experts appear due to incomplete information available regarding x. Therefore, precise estimation of its both membership value and non-membership value becomes highly difficult. Hence, there is a high chance that both membership value and the non-membership value assigned to x by the expert may not be absolutely correct. So, whenever we try to measure similarity between the IFSs using the various distance measures involving all the components of IFSs like membership value, non-membership value together with hesitation, then we often notice that all of them fails to describe the underlying situation completely. Therefore, the similarity measures derived from these distance measures also fails to produce good results. So, we introduce a new similarity measure by properly defining a similarity degree through the result established in this paper. The similarity measure has a central role in developing a modified λ-cutting algorithm for clustering. Here we also establish the efficacy of our modified λ-cutting algorithm while implementing it on a real world data set.

Reliability is the measure of the result of the quality of the system over a long run. The reliability-redundancy allocation problem (RRAP) aims to ensure high systems reliability in the presence of optimally redundant systems components. This is one of the most important design considerations for the systems designers. Several researchers have addressed this important issue during last few decades. However, due to the embedded uncertainty in the parameters of the system components, reliability as well as the costs of the whole system fits very well to be modeled as fuzzy quantity. We therefore modeled this problem as a fuzzy multi-objective optimization problem (MORRAP) that is addressed using the popular multi-objective evolutionary algorithm, NSGA-II (non-dominated sorting genetic algorithm-II). We have considered the based MORRAP with fuzzy type-2 uncertainty. As far as we know, no research has been reported where MORRAP was considered under type-2 fuzzy uncertainty. A typical numerical example is included and results are compared showing that our approach outperforms other recently reported results.

A computing with words/Per-C based user feedback collection model for controlling the processor power efficiency is introduced. Needless to say that CWW/Per-C is a very efficient tool in modelling human perceptions. Here the objects of computation are the words drawn from natural language instead of numbers [22], [23]. Perceptions alone don't make the sole criteria rather the backed logic of reasoning is also a supportive tool in the same scenario. In our present work, we have proposed a new algorithm viz. UFOPeC (user feedback optimized perceptual computing) for obtaining the optimal power efficiency in adaptive computing systems that can run at multiple operating voltages. Our approach models the user satisfaction very well and more realistically as compared to than the other existing mechanisms like HAPPE [1] as we have taken the user feedback in terms of words and modelled the same using the IT2 FSs (interval type-2 fuzzy sets). An appropriate numerical example has been chosen to demonstrate the design of our model.

The timing constraint of tasks in the mobile real-time computing systems plays the central role in deciding the task schedule as timely completion of the task is very important in such systems. These timing constraints are however completely unquantifiable during the time of system modeling and designing. Thus we consider type-2 fuzzy sets for modeling the timing constraints in mobile and time-critical computing systems and propose a new algorithm FT2EDF (Fuzzy Type-2 Earliest Deadline First) for task scheduling. On the other hand, because of the limitation of the storage power, power efficiency is another foremost design objective for designing mobile real-time computing systems. However, reduction of processor power pulls down the system performance. Timely task completion and power efficiency are therefore two mutually conflicting criteria. In this paper, we propose a heuristic based solution approach that with a modified version of the non-dominated sorting genetic algorithm-II (NSGA-II). Our approach allows that a processor dynamically switches between different voltage levels to ensure optimum reduction in the power requirements without compromising the timeliness of the task completion. The efficacy of our approach is demonstrated with two numerical examples. Comparison with the previous results show that our solution ensures approximately 44% of energy saving as compared to the around 25% of the earlier results.

In real-time systems, energy efficiency is a vital issue since usually such systems run on battery and are remotely placed. Another important aspect of these systems is their capabilities to produce timely results. In this paper, we have reported how these two conflicting issues of embedded real-time systems can be addressed with the help of an efficient evolutionary algorithm viz. NSGA-II (Non-Dominated Sorting Algorithm-II). Moreover, during the system design time, the timing parameters in real-time systems are all designers' approximation since those can hardly be predicted before runtime. This means that there exists some uncertainty and hence it is appropriate to consider fuzzy numbers to model these timing parameters. Although type-I fuzzy numbers were used by a number of researchers to model the timing parameters of real-time embedded systems, they suffer from the interpretability issues. To address this, we thus propose here to consider type-2 fuzzy numbers to model real-time tasks timing parameters. Few numerical examples are included to demonstrate our proposed technique.