Accurately monitoring sleep in an intelligent environment, such as a smart home, using a non-invasive system, is challenging. The challenge here is the real-world sleep component, particularly the blanket. The blanket plays a crucial role in sleep posture monitoring, influencing the thermal signature and, consequently, the classification accuracy. This paper presents a novel computational solution, the Cascaded Posture Classification Network (CPCN), designed to address the critical issue of thermal attenuation resulting from bedding conditions. The CPCN uses a dual-stage intelligent system that processes thermal images captured by an unobtrusive thermal sensor array. The system first employs a dedicated Blanket Classifier Convolutional Neural Network. It first analyses the bedding condition (No Blanket, Medium Blanket, Thick Blanket). The classification and routing at this stage achieved an accuracy of 94.44 %. The image is then routed to one of three specialised CNNs based on the blanket condition that developed in response to the thermal challenge. The four sleep postures (Prone, Supine, Left, Right) for the blanket conditions (No Blanket, Medium Blanket, Thick Blanket) achieved accuracies of$98.25 \%, 84 \%$, and 77 %, respectively. The system achieved an accuracy of 81.64 % across both stages. This novel cascaded design establishes a necessary methodological phase, moving thermal monitoring from laboratory settings to reliable, robust, and applicable real-home environments.

Online grooming is a significant and widely documented issue affecting children on social media platforms. These predatory interactions can inflict severe psychological harm on victims and, in tragic cases, have resulted in the death of a child. There is some literature on the clustering methods to determine behavioral insights in Online Grooming attacks and some classification methods considering Online Grooming beyond binary classification. However the aim of this paper is to extend this further by identifying if a transcript contains the transmission of sexual media and/or the arrangement of a physical meet up. The three semantic classifications using Bidirectional Encoder Representations from Transformers (BERT) returned 99.7% relevant insight clusters, with core insight F1-scores achieving 0.98 for sexual semantic (S), 0.89 for meet up (U), and 0.85 for media transmission (M), making up the ‘SUM’ approach. The key insight groupings of Sexual Media Transmission (SM) achieved an F1-score of 0.53. It is suggested that these key insight grouping scores can be improved by adjusting Drop Length and Step Values, to return differing sets of insight clusters. Further work consists of modifying this approach to differentiate between Fantasy Child Sex Offenders (FCSOs) and Contact Child Sex Offenders (CCSOs).

Due to their ability to provide radically enhanced security and user convenience, biometrics are becoming increasingly important as digitisation permeates everyday life. A myriad of biometric modalities, ranging from security systems to personal authentication technologies in mobile devices, has been successfully deployed across a wide range of applications. While modern biometric solutions leverage the latest advancements in AI, cybersecurity threats and attacks are also exploiting these advancements to become more sophisticated. To address this dual challenge, new biometric solutions must balance robustness with cost-effectiveness while preserving privacy. Focusing on visionbased gait biometrics, and building on our recent success in gait anomaly detection, which utilised robust, low-cost, low-resolution far-infrared thermal sensor arrays combined with deep learning autoencoders, we propose investigating this technology as a novel behavioural biometric modality. Specifically, we aim to explore its application in access control through personal authentication, focusing on enhancing home security. This research seeks to advance gait biometric systems by offering an AI-based complementary biometric, as an affordable additional layer of security that requires minimal training and computational resources while maintaining user privacy.

Heart rate variability serves as a valuable indicator and biomarker for stress detection and monitoring. Feature selection, which aims to identify relevant features from a large set of variables, is a crucial preprocessing step towards this. However, this task becomes challenging due to high dimensionality and the presence of irrelevant and redundant attributes. The Minimum Redundancy and Maximum Relevance (mRMR) feature selection method addresses this challenge by selecting relevant features while controlling redundancy. This paper presents extensions and evaluated versions of the mRMR feature selection methods for stress detection using Heart Rate Variability (HRV) measures. The proposed feature selection methods extend the traditional mRMR by replacing the Pearson correlation redundancy with non-linear feature redundancy measures capable of capturing more complex relationships between variables. An extensive empirical evaluation is conducted on the proposed mRMR extensions, comparing them with four other baseline feature selection methods using three publicly available datasets. The experimental results demonstrate the effectiveness of incorporating the non-linear feature redundancy measure into the feature selection process.

Customer service is an important and expensive aspect of business, often being the largest department in most companies. With growing societal acceptance and increasing cost efficiency due to mass production, service robots are beginning to cross from the industrial domain to the social domain. Currently, customer service robots tend to be digital and emulate social interactions through on-screen text, but state-of-the-art research points towards physical robots soon providing customer service in person. This article explores the feasibility of Transfer Learning different customer service domains to improve chatbot models. In our proposed approach, transfer learning-based chatbot models are initially assigned to learn one domain from an initial random weight distribution. Each model is then tasked with learning another domain by transferring knowledge from the previous domains. To evaluate our approach, a range of 19 companies from domains such as e-Commerce, telecommunications, and technology are selected through social interaction with X (formerly Twitter) customer support accounts. The results show that the majority of models are improved when transferring knowledge from at least one other domain, particularly those more data-scarce than others. General language transfer learning is observed, as well as higher-level transfer of similar domain knowledge. For each of the 19 domains, the Wilcoxon signed-rank test suggests that 16 have statistically significant distributions between transfer and non-transfer learning. Finally, feasibility is explored for the deployment of chatbot models to physical robot platforms including “Pepper”, a semi-humanoid robot manufactured by SoftBank Robotics, and “Temi”, a personal assistant robot.

Forgery of a signature with the aim of deception is a serious crime. Machine learning is often employed to detect real and forged signatures. In this study, we present results which argue that robotic arms and generative models can overcome these systems and mount false-acceptance attacks. Convolutional neural networks and data augmentation strategies are tuned, producing a model of 87.12% accuracy for the verification of 2,640 human signatures. Two approaches are used to successfully attack the model with false-acceptance of forgeries. Robotic arms (Line-us and iDraw) physically copy real signatures on paper, and a conditional Generative Adversarial Network (GAN) is trained to generate signatures based on the binary class of ‘genuine’ and ‘forged’. The 87.12% error margin is overcome by all approaches; prevalence of successful attacks is 32% for iDraw 2.0, 24% for Line-us, and 40% for the GAN. Fine-tuning with examples show that false-acceptance is preventable. We find attack success reduced by 24% for iDraw, 12% for Line-us, and 36% for the GAN. Results show exclusive behaviours between human and robotic forgers, suggesting training wholly on human forgeries can be attacked by robots, thus we argue in favour of fine-tuning systems with robotic forgeries to reduce their prevalence.

Abstract Human distance estimation is essential in many vital applications, specifically, in human localisation-based systems, such as independent living for older adults applications, and making places safe through preventing the transmission of contagious diseases through social distancing alert systems. Previous approaches to estimate the distance between a reference sensing device and human subject relied on visual or high-resolution thermal cameras. However, regular visual cameras have serious concerns about people’s privacy in indoor environments, and high-resolution thermal cameras are costly. This paper proposes a novel approach to estimate the distance for indoor human-centred applications using a low-resolution thermal sensor array. The proposed system presents a discrete and adaptive sensor placement continuous distance estimators using classification techniques and artificial neural network, respectively. It also proposes a real-time distance-based field of view classification through a novel image-based feature. Besides, the paper proposes a transfer application to the proposed continuous distance estimator to measure human height. The proposed approach is evaluated in different indoor environments, sensor placements with different participants. This paper shows a median overall error of $$\pm 0.2$$ ± 0.2 m in continuous-based estimation and $$96.8\%$$ 96.8 % achieved-accuracy in discrete distance estimation.

To support the independent living of older adults in their own homes, it is essential to identify their abnormal behaviors before triggering an automated alert system. Existing normal vision sensing approaches to detect human falls in the activities of daily living (ADL) experienced acceptability issues due to outstanding privacy concerns when they are deployed in personal environments. Besides, false alerts (false-positive) fall detection has not been addressed thoroughly in systems that report abnormal human behaviors as emergency alerts to the information support. This article proposes a novel human-in-the-loop fall detection approach in the ADLs using a low-resolution thermal sensor array. The motivation for enabling a human interactive model, fall detection confirmation, is to influence resource efficiency by reducing false-positive alerts while keeping the false-negative fall predictions as low as possible. The proposed approach is based on the motion sequence classification of human movements using a recurrent neural network. The proposed approach is evaluated with comprehensive experiments using different learning techniques, users, and domestic environment conditions. This article shows a performance accuracy of 99.7% to detect human falls from various typical ADLs.

Mobile applications in the area of human-centered applications are based on fuzzy logic have exhibited their effectiveness in managing intelligent environments, however the deployment of mobile fuzzy logic systems has been usually associated with dedicated hardware and software packages. Introducing openness for fuzzy logic systems offers exciting features such as system independence, simplicity, load balancing, and controlled resource allocation. On the other hand, while major cloud service providers support readymade commercial services for AI techniques such as for deep neural networks, there is no similar services for fuzzy logic systems. This study aims to develop a cloud-based fuzzy logic system under Microsoft Azure, employing Simpful as the cloud-side Python library and FML as data exchange standard. The developed cloud service is shown to effectively serve mobile phone applications for human monitoring purposes. Also in the present study, two types of fuzzy inference systems namely Mamdani and TSK have been utilized wherein both these systems have been compared on the basis of their processing time and accuracy of result. Results indicated that Mamdani fuzzy inference system outperformed TSK fuzzy inference system in terms of processing time by 0.456 seconds. Moreover, the detection accuracy of Mamdani system was found to be higher than that of TSK system by 6.82%.

The role of machine learning in neuroscience has been increasing through the years, in aiding diagnosis, biomarker discovery, signal analysis, and other applications. However, the lack of information of the decision-making of the models restricts their use and adoption by the community. In the process of neuronal signal acquisition, other electrical signals can distort the recording, for which a review process is necessary. Machine learning can aid by automatically detecting affected segments, speeding up the review process. However, as the ground-truth labelling is done manually or via a threshold, researchers must be able to identify the causes of false negatives and positives. This paper looks into explainable machine learning for artefact detection in invasively recorded neural signals through the use of different classifiers, trained with a feature subset produced by the combination of feature selection algorithms to reduce the dimensionality by two orders of magnitude. Our results show that the bagging decision tree model is best suited for creating a generalised model that is capable of classifying artefactual patterns in a multi-state dataset, which achieves an accuracy of 96.1%. Lastly, the predictor importance, Shapely values, and reduced feature space visualisation are used to gain insight into the model.

Human-centric applications of a single thermal sensor array (TSA) have performed extremely well in many areas. However, most of these works have not yet reached the real applicability stage of the Internet of Things (IoT) applications. The main limitation of deploying such systems on a large scale is the challenge of fusing multiple TSAs to cover a wide inspection area, e.g., smart homes, hospitals, and many other domestic environments. On the other hand, objects that appear in the low-resolution thermal images acquired from TSA have low intraclass variations and high interclass similarities, making the identification of the overlapping regions through matching a comparable template image in multiple images very difficult. This article proposes a motion-based approach to fuse multiple TSAs and learn the domestic environment layout to enable further human-centred IoT applications to run in the cloud. Besides, a privacy improvement on utilizing these sensors in IoT applications is proposed. The proposed approach is evaluated with comprehensive experiments on different sensor placements and domestic environment conditions. This article shows an average performance of 92.5% accuracy using various machine learning techniques and use case scenarios.

Fuzzy logic systems are customarily related to specific hardware or software systems. Nevertheless, it has been observed that distributed and cloud-based architectures of various intelligent systems are pouring intensifying attention. While the distributed architectures can potentially add values in developing fuzzy systems, a lack of standard methods and practices may limit their public use. This study aims to provide a standard solution for developing cloud-based service-oriented architectures for fuzzy logic systems, based on extending IEEE-1855 (2016) in the defining system and exchanging data. Experiments were performed employing simulation concerning collection, processing and monitoring of data in a distributed manner over the web. A real-time human activity recognition simulated scenario is also demonstrated through a cloud-based fuzzy system.

Multi-agent multi-target search problems, where the targets are capable of movement, require sophisticated algorithms for near-optimal performance. While there are several algorithms for agent control, comparatively less attention has been paid to near-optimal target behaviours. Here, a state-of-the-art algorithm for targets to avoid a single agent called TrailMax has been adapted to work within a multiple agents and multiple targets framework. The aim of the presented algorithm is to make the targets avoid capture as long as possible, if possible until timeout. Empirical analysis is performed on grid-based gaming benchmarks. The results suggest that Multiple Pursuers TrailMax reduces the agent success rate by up to 15% as compared to several previously used target control algorithms and increases the time until capture in successful runs.

The improvement of hardware for the acquisition and processing of electroencephalography (EEG) has made its portability become a reality. This allows for studies to be carried outside lab settings, as well as many commercial applications. As recordings are done over extended periods, these devices generate large volumes of data, mainly if the neuronal activity is recorded through multiple channels. Machine learning (ML) techniques allow to effectively analyse and use this data for a wide range of applications. However the portability of these techniques can be challenging. In this article, we set out to review over 40 relevant articles where ML techniques in a diverse set of EEG applications that have successfully been incorporated into portable systems.

Recent advancements in the Internet of Things and the miniaturisation of low-cost sensing devices allow for the unobtrusive collection of data for human activity recognition and behaviour modelling. A useful application of this in the context of ambient assisted living is in the monitoring of older adults daily for improved wellbeing and quality of life. The existing solutions are based on per-individual monitoring, therefore the systems are managed independently for each ambient intelligent environment. In this paper, we proposed a centralised system for the collective monitoring of individuals in a community. The proposed approach is based on a cloud-based solution where data collection and processing are centralised. Since the data are aggregated for all the residents, the system has the potential of promoting social interaction among the community residents. Additionally, the cost of the in-home monitoring system can be reduced since only the sensing devices are required for data collection, while the processing is carried out on the cloud infrastructure. This also reduces the tedious tasks required in setting up individual home monitoring systems. The role of assistive robots, the possibility of remote monitoring and potential challenges of the proposed approach are explored.

Falls are considered as one of the greatest risks and a fundamental problem in health-care for older adults living alone at home. The number of older adults living alone in their own homes is increasing worldwide due to the high expense of health care services. Therefore, it is important to develop an accurate system with the ability to detect human falls during daily activities. The focus of this study is to distinguish and detect human falls in Activities of Daily Living (ADL) based on data acquired from an accelerometer device. In this paper, a novel method based on Fuzzy Entropy measure is investigated to detect and distinguish human fall from other activities with a high degree of accuracy. The proposed method is tested and evaluated based on a publicly available URFD dataset. The experimental results show that Fuzzy Entropy achieved a sensitivity and specificity of 100% and 97.8%, respectively. Comparisons with other methods have also provided further support to the proposed method.

Human activity recognition is concerned with identifying the specific movement of a person based on sensor data. In recent years, many different techniques have been proposed for modelling and recognising human activities, with a specific focus on the development of approaches to classifying human activities using deep learning techniques. The research presented in this paper proposes a fuzzy feature representation approach to represent occupancy sensor data, along with a Convolutional Neural Network Classifier (CNNC) for human activity modelling and recognition. Sensory data gathered from a home environment are converted into occupancy data representing human activities and then fuzzified before being fed as inputs into the CNNC. The learning capability of CNNC allows the model to learn the relationship between the fuzzified inputs and their corresponding output activities during training mode. The relations learned in the trained CNNC model are then used to identify human activity patterns and classify these when the testing dataset is applied. The proposed method is evaluated using a dataset representing activities of daily living for a single user gathered from a real-home environment.

There is a growing demand for efficient and privacy-preserving intelligent solutions in a multi-occupancy environment. This paper proposes a non-contact scheme for occupancy estimation using an infrared thermal sensor array, which has the advantages of low-cost, low-power, and high-performance capabilities. The proposed scheme offers an accurate human heat segmentation technique that extracts human body temperature from a noisy environment. It is shown that the proposed system can detect the empty occupancy state after utilising the segmentation technique with an accuracy of 100%. By using adaptive boosting, it is shown that the system is capable of measuring the non-empty occupancy with an overall accuracy of 98.2%.

Fuzzy Logic Systems (FLSs) have shown their potentials in Ambient Intelligence (AmI) applications. However, the implementation of FLSs is typically linked to dedicated and non-scalable hardware/software systems. As a result, some specific AmI requirements such as web communications and Service-Oriented Architecture (SOA), which can be found in many modern systems, are rarely adapted for FLSs. Sharing FLSs accessibility as web services (called fuzzy-as-a-service), in which the service is developed independently from a specific FLS, allows for system autonomy, openness, load balancing, efficient resource allocation and eventually cost-efficiency, particularly for computationally intense FLSs. In a wider context, such features can open new dimensions for FLSs' applicability in Cloud Computing and Internet of Things devices. Recent advances in standardising Fuzzy Mark-up Language (IEEE 1855-2016) and its associated software libraries (such as JFML) has made this even more achievable. This paper proposed fuzzy-as-a-service architecture based on IEEE 1855-2016, JFML and SOA. Through a simulated experiment, this paper concerned the collection, processing and monitoring the distributed data over the web, thus a real-time human activity recognition simulated scenario using a rule based FLS is demonstrated.

The neural recordings known as Local Field Potentials (LFPs) provide important information on how neural circuits operate and relate. Due to the involvement of complex electronic apparatuses in the recording setups, these signals are often significantly contaminated by artifacts generated by a number of internal and external sources. To make the best use of these signals, it is imperative to detect and remove the artifacts from these signals. Hence, this work proposes a pattern recognition neural network based single-channel automatic artifact detection tool. The tool is capable of detecting the artifacts with an 93.2% of overall accuracy and requires an average computing time of 2.57 seconds to analyse LFPs of one minute duration, making it a strong candidate for online deployment without the need for employing high performance computing equipment.

Cloud of Things (CoT) is increasingly viewed as a paradigm that can satisfy the diverse requirements of emerging Internet of Things (IoT) applications. The potential of CoT is not yet realized due to challenges in sharing and reusing IoT physical resources across multiple applications. The existing approaches provide small-scale and hardware-dependent shared access to IoT resources. This article considers using market mechanisms to commoditize CoT resources as the approach to enable shared access to CoT resources and to improve their reusability. In order to achieve this, the requirements for trading CoT resources are discussed to conceptualize the proposed approach. A generic description model for CoT resource is introduced to quantify the value of CoT resources. In this article, a marketplace architecture for trading CoT resources referred to as AMACoT is proposed. By formulating the trading of CoT resources as an optimization problem, the proposed approach is experimentally validated. The evaluation measures the system performance and verifies the optimization problem using three evolutionary algorithms. The evaluation of the optimization algorithms demonstrates the optimality of trading CoT resources solutions in terms of resource cost, resource utilization, provider lock-in, and provider profit.

The emerging Cloud of Things (CoT) paradigm promises to meet the diverse requirements of many real-world applications, which previously could not be fulfilled by either cloud computing or Internet of Things (IoT). Trading CoT resources is a challenging aspect, particularly when managing quality of service (QoS) as resource providers and application developers have different priorities. This article focuses on the challenge of supporting QoS when trading CoT resources and performing resource allocation. The contributions of this article are: 1) the problem of managing QoS while trading CoT resources is investigated as an optimization problem; 2) a QoS model is proposed to solve the problem by optimizing five different QoS objectives; and 3) experimental evaluation of the proposed model using three optimization algorithms. The evaluation results show the efficiency and dynamism of the proposed model in optimizing CoT resource allocation based on diverse QoS objectives, including resource cost, energy consumption, response time, fault tolerance, and resource coverage.

Assistive robots in ambient assisted living environments can be equipped with learning capabilities to effectively learn and execute human activities. This paper proposes a human activity learning (HAL) system for application in assistive robotics. An RGB-depth sensor is used to acquire information of human activities, and a set of statistical, spatial and temporal features for encoding key aspects of human activities are extracted from the acquired information of human activities. Redundant features are removed and the relevant features used in the HAL model. An ensemble of three individual classifiers—support vector machines (SVMs), K -nearest neighbour and random forest—is employed to learn the activities. The performance of the proposed system is improved when compared with the performance of other methods using a single classifier. This approach is evaluated on experimental dataset created for this work and also on a benchmark dataset—the Cornell Activity Dataset (CAD-60). Experimental results show the overall performance achieved by the proposed system is comparable to the state of the art and has the potential to benefit applications in assistive robots for reducing the time spent in learning activities.

In this paper we apply the Fuzzy Entropy and Approximate Entropy measures to the Activities of Daily Living (ADL) for a set of elderly subjects in their own homes, and compare the entropy measures against a simpler count of activity transitions. The aim is to assess whether a single relatively simple measure can give an overview of the ADL in order to provide summaries of the wellbeing of an elderly person for their carers or relatives. We find that both the entropy measures and activity count vary considerably between different elderly people, and from day to day. The absolute level of these measures seems to be indicative of different ADL levels. However, it appears that further analysis over a longer period and with more annotation by the volunteers is necessary before the measures can be appropriately interpreted.

In this article, a hybrid technique for user activities outliers detection is introduced. The hybrid technique consists of a two‐stage integration of principal component analysis and fuzzy rule‐based systems. In the first stage, the Hamming distance is used to measure the differences between different activities. Principal component analysis is then applied to the distance measures to find two indices of Hotelling'sT2and squared prediction error. In the second stage of the process, the calculated indices are provided as inputs to the fuzzy rule‐based systems to model them heuristically. The model is used to identify the outliers and classify them. The proposed system is tested in real home environments, equipped with appropriate sensory devices, to identify outliers in the activities of daily living of the user. Three case studies are reported to demonstrate the effectiveness of the proposed system. The proposed system successfully identifies the outliers in activities distinguishing between the normal and abnormal behavioral patterns.

This paper presents a system for automatically classifying the resting location of a moving object in an indoor environment. The system uses an unsupervised neural network (Self Organising Feature Map) fully implemented on a low-cost, low-power automated home-based surveillance system, capable of monitoring activity level of elders living alone independently. The proposed system runs on an embedded platform with a specialised ceiling-mounted video sensor for intelligent activity monitoring. The system has the ability to learn resting locations, to measure overall activity levels and to detect specific events such as potential falls. First order motion information, including first order moving average smoothing, is generated from the 2D image coordinates (trajectories). A novel edge-based object detection algorithm capable of running at a reasonable speed on the embedded platform has been developed. The classification is dynamic and achieved in real-time. The dynamic classifier is achieved using a SOFM and a probabilistic model. Experimental results show less than 20% classification error, showing the robustness of our approach over others in literature with minimal power consumption. The head location of the subject is also estimated by a novel approach capable of running on any resource limited platform with power constraints.

Analysis of the office workers' activities of daily working in an intelligent office environment can be used to optimize energy consumption and also office workers' comfort. To achieve this end, it is essential to recognise office workers' activities including short breaks, meetings and non-computer activities to allow an optimum control strategy to be implemented. In this paper, fuzzy finite state machines are used to model an office worker's behaviour. The model will incorporate sensory data collected from the environment as the input and some pre-defined fuzzy states are used to develop the model. Experimental results are presented to illustrate the effectiveness of this approach. The activity models of different individual workers as inferred from the sensory devices can be distinguished. However, further investigation is required to create a more complete model.

In this paper we present an ambient intelligence system for modelling and control of power use within an office environment. We define a multi-scale model of the office worker, consisting of a coarse grained office user profile, together with a fine grained characteristics model to summarise their behaviours. Sensor data gathered from individual offices are used to create these models. Collected data are fuzzified to produce meaningful and understandable descriptions of the worker's daily activities. These models of the individuals form the basis of a Dynamic Power Usage Scheme (DPUS) to control individual offices. A trial version of the system has been implemented using individual academic staff offices in a university campus. Experiments have shown that even a simplistic user profile can lead to significant power savings from appropriate control of computers in an office environment.

In today's troubled economies, nations are competing in many aspects, including innovation and knowledge progress. Even though there are many composite indicators to measure knowledge and innovation at both micro and macro levels, benefits to decision makers still limited due to numerous progress indicators, without any unified, easy to visualize and evaluate forecasting capabilities. This paper introduces a novel approach to forecasting and finding the aggregated position of many Knowledge-Based Economy (KBE) with a high degree of accuracy. The suggested approach is based on data mining, Neural Networks, Principle Component Analysis (PCA), and Self-Organising Map (SOM). The proposed model has the capability of forecasting and aggregating five major KBE indicators into a unified meaningful map that places any KBE in its league regardless of incomplete missing or little data. The Unified Knowledge Economy Forecast Map (UKFM) reflects the overall position of homogeneous knowledge economies, and it can be used to visualise, identify or evaluate stable, progressing or accelerating KBEs.

The aim of this paper is to design a unified knowledge economy competitiveness index using fuzzy clustering, to aggregate four of the most reputable and famous knowledge economy indicators into a unified index that reflects the overall rate of knowledge in an economy, to serve many purposes for the decision makers and foreign investors interested in such economy. The four selected indices are: Knowledge Economy Index (KEI) from World Bank, Information and Communication Technologies Development Index (IDI) from United Nations agency for information and communication technology issues (ITU), Global Competitiveness Index (GCI) from the World Economic Forum, and World Competitiveness Yearbook (WCY) from Institute for Management Development (IMD). To achieve this unified index, a four steps framework is proposed. The first step utilizes a Correlation analysis, the second step is to carry a Principle Component Analysis (PCA) analysis and the third step employs training an Adaptive Neural Fuzzy Inference Systems (ANFIS) and the forth step is to create a unified index based on all existing indices. The purpose of the first step was to test the relationship between the selected indices and how strong it is. The PCA is employed to test the similarity amongst existing indices and whether they can be reduced in any form. ANFIS was used to generate rules to create trained submodel that determine which of the input indices make efficient contribution to the new unified knowledge indicator. Then, the fuzzy c-means clustering technique is used to construct the new Unified Knowledge Competitiveness and Progress Indicator (UKPI) which combines the four selected aggregate indices into a new single meaningful index that reflects the overall rate of Knowledge competitiveness and progress in a nation.

In this paper, the simulation of the environmental conditions in university offices is addressed. Based on the real data collected from the environment, a flexible simulator is developed which accepts different office worker pro- files including expected office occupancy and computer usage. The simulator generates sensory signals which represents different activities and occupancy of the office environment. These signals will then be used to control heating, lighting and computer stand-by mode of operation. The validity of the simulator is verified by tuning the simulator parameters to occupancy data collected by sensory systems from real offices.

The aim of this paper is to examine the suitability of binary similarity and dissimilarity measures in identifying frequent and abnormal human behavioural patterns in a smart home. There has been an increasing interest in this subject to help the elderly and disabled people to live alone in their own homes with little help and support from their carer. Similarity and dissimilarity measures have been applied for a wide range of pattern recognition tasks such as character recognition, image retrieval, etc. In this work, the binary similarity and dissimilarity indices are used on data generated from occupancy sensors including door and motion sensors in a smart home. These sensors indicate the presence and absence of the occupant in a specific area in the home. Many measures are introduced in the literature, the focus in this paper is on the measures that give credits to both the positive matches and the mismatching between two sensor values. This paper first gives an overview of the similarity measures to find the most common patterns and then dissimilarity or distance measures are used to identify unexpected or abnormal data. Data from two different case studies are used to validate the accuracy of these measurements.

In this paper, tagging techniques are employed to identify monitored inhabitants in smart environments. The aim of the work is to recognise the presence of tagged inhabitants in different areas of the environment. RSSI - based wireless localising sensory agents are used in separate areas to measure the distance of the tagged persons from the reader agents. Then, clustering methods are applied to the radio signal strength from the reader agents to find the data cluster of each area representing the occupied area by the tagged inhabitants. In this paper two different technologies including active RFID and ZigBee wireless technology are compared in distance-based measurements and an area-based experiment. It is shown that despite the existing uncertainty in radio signal strength, deploying multiple reader agents and applying a regional clustering can substantially reduce the uncertainty for area occupancy detection purpose.

In this paper, occupancy pattern extraction and prediction in an intelligent inhabited environment is addressed. The results of this research will help elderly people to live independently in their own home longer and help them in case of an emergency. Using a wireless sensor network system, daily behavioral patterns of the occupant are extracted. This information is then used to build a behavioral model of the occupant which ultimately is used to predict the future values representing the expected occupancy and other activities. The occupancy signal is represented by a long sequence of binary series indicating presence or absence of the occupant in a specific area. It is essential to convert this series of binary data into a more flexible and efficient format before it is applied for any further analysis and prediction. After converting the occupancy binary signals, the prediction model is built through a recurrent dynamic network, with feedback connections enclosing several layers of a Nonlinear Autoregressive netwoRk with eXogenous inputs (NARX) network. The results reported here shows that NARX provide better prediction results than conventional recurrent neural networks such as Elman networks. The case study reported here is based on a one bedroom flat with a single occupant.

Pattern analysis and prediction of sensory data is becoming an increasing scientific challenge and a massive economical interest supports the need for better pattern mining techniques. The aim of this paper is to investigate efficient mining of useful information from a sensor network representing an ambient intelligence environment. The goal is to extract and predict behavioral patterns of a person in his/her daily activities by analyzing the time series data representing the behaviour of the occupant, generated using occupancy sensors. There are various techniques available for analysis and prediction of a continuous time series signal. However, the occupancy signal is represented by a binary time series where only discrete values of a signal are available. To build the prediction model, recurrent neural networks are investigated. They are proven to be useful tools to solve the difficulties of the temporal relationships of inputs between observations at different time steps, by maintaining internal states that have memory. In this paper, a special form of recurrent neural network, the so-called Echo State Network (ESN) is used in which discrete values of time series can be well processed. Then, a model developed based on ESN is compared with the most popular recurrent neural networks; namely Back Propagation Through Time (BPTT) and Real Time Recurrent Learning (RTRL). The results showed that ESN provides better prediction results compared with BPTT and RTRL. Using ESN, large datasets are learnt in only few minutes or even seconds. It can be concluded that ESN are efficient and valuable tools in binary time series prediction. The results presented in this paper are based on simulated data generated from a simulator representing a person in a 1 bed room flat.

In this paper, the prediction of the occupancy of different areas in a single-occupant intelligent inhabited environment is addressed It is aimed to deliver a well-being monitoring and assistive environment to support elderly to live independently. A wireless sensor network of motion detection sensors is constructed to collect the required occupancy data. Individual sensory data are combined to form an occupancy time series. Then, a fuzzy predictor model is proposed to model the occupancy time series and the results of this technique are compared with other techniques in time series prediction including auto regressive moving average, adaptive network based fuzzy inference system, as well as transductive neuro-fuzzy inference model with weighted data normalization.

In this paper, a review of prediction techniques suitable for ambient intelligence environments is presented. Prediction challenges in sensor networks are considered in two phases including pattern extraction and rule matching. The prediction techniques reviewed in this paper come from two main research areas, namely, data mining and soft computing techniques. Moreover, a statistical modelling technique based on Markov chain is also considered. In this paper, we identify the centralized and distributed techniques of both data mining and soft computing areas. In addition, we identify the distributed approaches that utilize computational power of sensors in an ambient intelligence environment. Moreover, we show that some techniques use compression, regression or fuzzy methods to reduce the size of the collected sensory data.

A novel 9 rules self-tuning PD+I fuzzy logic controller applicable for a class of nonlinear plants is proposed in this paper. The controller comprises of three separate fuzzy logic controllers with each uses minimum number of rules and the output scaling factor is tuned automatically depending on the tracking error dynamic conditions. The controller is applied to a two-link revolute robot for the tracking control. Simulation results show that the robustness and tracking performance of the proposed controller is comparable to standard PD+I fuzzy logic controller at low and medium speed motions. However, the performance of the proposed new design far exceeds the standard design at high speed motions.

In this paper, a fuzzy supervisory indirect learning predictive controller (FsiLPC) is proposed. The controller integrates the concepts of the conventional model based predictive control (MBPC), the controller output error method and the fuzzy rule based system. In contrast to the conventional MBPC, the design of FsiLPC is more generic in terms of model compatibility; the predictive model to be adopted in the scheme can be of any structure. The FsiLPC demonstrates a number of attractive characteristics for industrial implementation. These include flexibility of employing a predictive model, adaptive capability to the variation in the operating conditions, and simplicity of the operating mechanisms. The methodology is thus considered as a potential solution to the control of complex multi-input-multi-output (MIMO) systems. The strategy of the integration is first illustrated through an overview of the general architecture of the controller, followed by descriptions of the individual operating mechanism. Two case studies are investigated to evaluate the performances of the controller. The first case study describes the control of a simple system with non-minimum phase dynamics while the control of a multivariable industrial process of a single screw extruder is detailed in the second case study. The encouraging results stem the interest to further explore the applicability of the proposed controller for industrial uses.

In the original paper of Jang and Sun (ibid., vol.4, p. 156-9, 1993), it is claimed that under a set of minor restrictions radial basis function networks and fuzzy inference systems are functionally equivalent. The paper shows that this set of restrictions is incomplete and that, when it is completed, the said functional equivalence applies only to a small range of fuzzy inference systems. In addition, a modified set of restrictions is proposed which is applicable for a much wider range of fuzzy inference systems. The original authors reply that they concerned themselves only with input-output functions, and not with the whole of learning. They also draw attention to the intention of their paper, written at a time when the field of study was much less mature.

The controller output error method (COEM) is introduced and applied to the design of adaptive fuzzy control systems. The method employs a gradient descent algorithm to minimize a cost function which is based on the error at the controller output. This contrasts with more conventional methods which use the error at the plant output. The cost function is minimized by adapting some or all of the parameters of the fuzzy controller. The proposed adaptive fuzzy controller is applied to the adaptive control of a nonlinear plant and is shown to be capable of providing good overall system performance.

In this paper, a matrix formulation of fuzzy rule based systems is introduced. A gradient descent training algorithm for the determination of the unknown parameters can also be expressed in a matrix form for various adaptive fuzzy networks. When converting a rule-based system to the proposed matrix formulation, only three sets of linear/nonlinear equations are required instead of set of rules and an inference mechanism. There are a number of advantages which the matrix formulation has compared with the linguistic approach. Firstly, it obviates the differences among the various architectures; and secondly, it is much easier to organize data in the implementation or simulation of the fuzzy system. The formulation will be illustrated by a number of examples.

An adaptive membership function scheme for general additive fuzzy systems is proposed in this paper. The proposed scheme can adapt a proper membership function for any nonlinear input-output mapping, based upon a minimum number of rules and an initial approximate membership function. This parameter adjustment procedure is performed by computing the error between the actual and the desired decision surface. Using the proposed adaptive scheme for fuzzy system, the number of rules can be minimized. Nonlinear function approximation and truck backer-upper control system are employed to demonstrate the viability of the proposed method.