Abstract Human Activity Recognition (HAR) covers methods for automatically identifying human activities from a stream of data. End‐users of HAR methods cover a range of sectors, including health, self‐care, amusement, safety and monitoring. In this survey, the authors provide a thorough overview of deep learning based and detailed analysis of work that was performed between 2018 and 2023 in a variety of fields related to HAR with a focus on device‐free solutions. It also presents the categorisation and taxonomy of the covered publication and an overview of publicly available datasets. To complete this review, the limitations of existing approaches and potential future research directions are discussed.

Discovering knowledge from social media becomes a trend in many domains such as tourism, where users' feedback and rating are the basis of recommendation systems. In this context, cluster analysis has been a major tool to disclose user groups by which the process of collaborative filtering can better determine a personalised suggestion. Matching this to the curse of big data is a challenge with previous studies either implementing conventional techniques on a distributed system or making use of data sampling. Specific to ensemble clustering, only a few aim to obtain both scalability and privacy preserving that are significant to handling social data. This paper presents a new bi-level framework of ensemble clustering in which an instance-segment based analysis is adopted to ensure data privacy and reduce the complexity of clustering the whole dataset. Unlike existing studies, instead of drawing a single clustering from each segment, multiple clusterings are selected to better represent instances therein. Based on published tourism datasets and different experimental settings, the new approach usually outperforms its baselines whilst being competitive to related methods found in the literature. Additional case studies on simulated big datasets and noisy variations are reported and discussed in addition to the analysis of algorithmic parameters.

Attack detection is one of the main features required in modern defence systems. Despite the ongoing research, it remains challenging for a typical mechanism like network-based intrusion detection system (NIDS) to catch up with evolving adversarial attacks. They specifically aim to confuse a machine-learning based predictor. Without the knowledge of adversarial patterns, the best approach is generalising signatures learned from a dataset of legitimate connections and known intrusions. This work focuses on analysing non-payload traffics so that the resulting techniques can be exploited to a range of network-based applications. It investigates a novel means to deal with the problem of imbalanced classes. An optimised undersampling method is introduced to select a subset of majority-class representatives initially created through an ensemble clustering procedure. A weighted combination of criteria representing distributions within and between classes is proposed as the objective function for a global optimisation using the artificial bee colony (ABC). This approach usually outperforms its baselines and other state-of-the-art undersampling models, with ABC being more effective using the global best strategy than a random selection of solutions or an iterative greedy search. The paper also details the parameter analysis offering a heuristic guide for potential taking up of the proposed techniques.

Malware has remained a consistent threat since its emergence, growing into a plethora of types and in large numbers. In recent years, numerous new malware variants have enabled the identification of new attack surfaces and vectors, and have become a major challenge to security experts, driving the enhancement and development of new malware analysis techniques to contain the contagion. One of the preliminary steps of malware analysis is to remove the abundance of counterfeit malware samples from the large collection of suspicious samples. This process assists in the management of man and machine resources effectively in the analysis of both unknown and likely malware samples. Hashing techniques are one of the fastest and efficient techniques for performing this preliminary analysis such as fuzzy hashing and import hashing. However, both hashing methods have their limitations and they may not be effective on their own, instead the combination of two distinctive methods may assist in improving the detection accuracy and overall performance of the analysis. This paper proposes a Fuzzy-Import hashing technique which is the combination of fuzzy hashing and import hashing to improve the detection accuracy and overall performance of malware analysis. This proposed Fuzzy-Import hashing offers several benefits which are demonstrated through the experimentation performed on the collected malware samples and compared against stand-alone techniques of fuzzy hashing and import hashing.

Cancer has been identified as the leading cause of death. It is predicted that around 20-26 million people will be diagnosed with cancer by 2020. With this alarming rate, there is an urgent need for a more effective methodology to understand, prevent and cure cancer. Micro array technology provides a useful basis of achieving this ultimate goal. In particular to cancer research, it has become almost routine to create gene expression profiles, which can discriminate patients into good and poor prognosis groups, and identify possible tumor subtypes. This classification or predictive model offers a useful tool for individualized treatment of disease. However, the accuracy of existing classifiers have been constrained by the curse of dimensionality typically observed in micro array data. In addition to gene selection, one may transform the original data to another variation, where only key gene components are included. Unlike conventional transformation-based techniques found in the literature, this paper presents a novel method that makes use of cluster ensembles, specifically the summarizing information matrix, as the transformed data for the following classification step. Among different state-of-the-art methods, the link-based cluster ensemble approach (LCE) provides a highly accurate clustering, and thus particularly employed here. The performance of this transformation model is evaluated on published micro array datasets and C4.5, in comparison with benchmark techniques. The findings suggest that the new model can improve the classification accuracy of original data and performs better than the other transformation methods investigated in the empirical study.

Cluster ensembles have been shown to be better than any standard clustering algorithm at improving accuracy. This meta-learning formalism helps users to overcome the dilemma of selecting an appropriate technique and the parameters for that technique, given a set of data. It has proven effective for many problem domains, especially microarray data analysis. Among different state-of-the-art methods, the link-based approach (LCE) recently introduced by [22], [23] provides a highly accurate clustering. This paper presents the improvement of LCE with a new link-based similarity measure being developed and engaged. Additional information that is already available in a network is included in the similarity assessment. As such, this refinement can increase the quality of the measures, hence the resulting cluster decision. The performance of this improved LCE is evaluated on synthetic and UCI benchmark datasets, in comparison with the original and several well-known cluster ensemble techniques. The findings suggest that the new model can improve the accuracy of LCE and performs better than the others investigated in the empirical study.

Cancer has been identified as the leading cause of death. It is predicted that around 20-26 million people will be diagnosed with cancer by 2020. As a result, there is an urgent need for a more effective methodology to prevent and cure cancer. Microarray technology provides a useful basis of achieving this ultimate goal. For cancer research, it has become almost routine to create gene expression profiles, which can discriminate patients into good and poor prognosis groups. This cluster analysis offers a useful basis for individualized treatment of disease. Cluster ensembles have been shown to be better than any standard clustering algorithm for such a task. This meta-learning formalism helps users to overcome the dilemma of selecting an appropriate technique and the parameters for that technique, given a set of data. Among different state-of-the-art methods, the link-based approach (LCE) provides a highly accurate clustering. This paper presents the improvement of LCE with a new link-based similarity measure being developed and engaged. Additional information that is already available in an information network is included in the similarity assessment. As such, this refinement can increase the quality of the measures, hence the resulting cluster decision. The performance of this improved LCE is evaluated on published microarray datasets, in comparison with the original LCE and several well-known cluster ensemble techniques. The findings suggest that the new model can improve the accuracy of LCE and performs better than the others investigated in the empirical study.

Gene expressions measured using microarrays usually encounter the problem of missing values. Leaving this unsolved may critically degrade the reliability of any consequent down-stream analysis or medical application. Yet, a further study of microarray data might be impossible with many analysis methods requiring a complete data set. This paper introduces a new methodology to impute missing values in microarray data. The proposed algorithm, CKNN impute, is an extension of k nearest neighbor imputation with local data clustering being incorporated for improved quality and efficiency. Gene expression data is typically represented as a matrix whose rows and columns correspond to genes and experiments, respectively. CKNN kicks off by finding a complete dataset via the removal of rows with missing value(s). Then, k clusters and their corresponding centroids are obtained by applying a clustering technique on the complete dataset. A set of similar genes of the target gene (with missing values) are those belonging to the cluster, whose centroid is the closest the target. Having known this, the target gene is imputed by applying k nearest neighbor method with similar genes previously determined. Empirical evaluation with published gene expression datasets suggest that the proposed technique performs better than the classical k nearest neighbor method and its extension found in the literature.

Although attempts have been made to solve the problem of clustering categorical data via cluster ensembles, with the results being competitive to conventional algorithms, it is observed that these techniques unfortunately generate a final data partition based on incomplete information. The underlying ensemble-information matrix presents only cluster-data point relations, with many entries being left unknown. The paper presents an analysis that suggests this problem degrades the quality of the clustering result, and it presents a new link-based approach, which improves the conventional matrix by discovering unknown entries through similarity between clusters in an ensemble. In particular, an efficient link-based algorithm is proposed for the underlying similarity assessment. Afterward, to obtain the final clustering result, a graph partitioning technique is applied to a weighted bipartite graph that is formulated from the refined matrix. Experimental results on multiple real data sets suggest that the proposed link-based method almost always outperforms both conventional clustering algorithms for categorical data and well-known cluster ensemble techniques.

Alias detection has been the significant subject being extensively studied for several domain applications, especially intelligence data analysis. Many preliminary methods rely on text-based measures, which are ineffective with false descriptions of terrorists' name, date-of-birth, and address. This barrier may be overcome through link information presented in relationships among objects of interests. Several numerical link-based similarity techniques have proven effective for identifying similar objects in the Internet and publication domains. However, as a result of exceptional cases with unduly high measure, these methods usually generate inaccurate similarity descriptions. Yet, they are either computationally inefficient or ineffective for alias detection with a single-property based model. This paper presents a novel orders-of-magnitude based similarity measure that integrates multiple link properties to refine the estimation process and derive semantic-rich similarity descriptions. The approach is based on order-of-magnitude reasoning with which the theory of fuzzy set is blended to provide quantitative semantics of descriptors and their unambiguous mathematical manipulation. With such explanatory formalism, analysts can validate the generated results and partly resolve the problem of false positives. It also allows coherent interpretation and communication within a decision-making group, using this computing-with-word capability. Its performance is evaluated over a terrorism-related data set, with further generalization over publication and email data collections.

Many approaches have been developed for academic performance evaluation using various fuzzy techniques. Initial methods rely greatly on experts' specification of analytical parameters, without making use of valuable information embedded in collected data. Given this insight, fuzzy rule induction has recently been studied as a data-driven alternative. Despite its efficiency and reported performance, the fuzzy subsethood metric representing the strength of relations between system variables is only used at a coarse level, with the underlying semantics being unfortunately distorted via a local re-scaling scheme. To overcome this problem, a novel fuzzy classification system is introduced in this paper, in which the existing measure is used to its full potential via the methodology of qualitative link analysis. With a network representation where variables and their relations are encoded as graph nodes and edges, the classification of a new instance conceptually becomes a problem of link-based similarity estimation that can be effectively resolved using the proposed fuzzy qualitative model. This new approach has been evaluated against the existing rule-based method, revealing significant advantages of the present work.

Cluster ensembles have recently emerged as a powerful alternative to standard cluster analysis, aggregating several input data clusterings to generate a single output clustering, with improved robustness and stability. From the early work, these techniques held great promise; however, most of them generate the final solution based on incomplete information of a cluster ensemble. The underlying ensemble-information matrix reflects only cluster-data point relations, while those among clusters are generally overlooked. This paper presents a new link-based approach to improve the conventional matrix. It achieves this using the similarity between clusters that are estimated from a link network model of the ensemble. In particular, three new link-based algorithms are proposed for the underlying similarity assessment. The final clustering result is generated from the refined matrix using two different consensus functions of feature-based and graph-based partitioning. This approach is the first to address and explicitly employ the relationship between input partitions, which has not been emphasized by recent studies of matrix refinement. The effectiveness of the link-based approach is empirically demonstrated over 10 data sets (synthetic and real) and three benchmark evaluation measures. The results suggest the new approach is able to efficiently extract information embedded in the input clusterings, and regularly illustrate higher clustering quality in comparison to several state-of-the-art techniques.

The measure of data reliability has recently proven useful for a number of data analysis tasks. This paper extends the underlying metric to a new problem of soft subspace clustering. The concept of subspace clustering has been increasingly recognized as an effective alternative to conventional algorithms (which search for clusters without differentiating the significance of different data attributes). While a large number of crisp subspace approaches have been proposed, only a handful of soft counterparts are developed with the common goal of acquiring the optimal cluster-specific dimension weights. Most soft subspace clustering methods work based on the exploitation of k-means and greatly rely on the iteratively disclosed cluster centers for the determination of local weights. Unlike such wrapper techniques, this paper presents a filter approach which is efficient and generally applicable to different types of clustering. Systematical experimental evaluations have been carried out over a collection of published gene expression data sets. The results demonstrate that the reliability-based methods generally enhance their corresponding baseline models and outperform several well-known subspace clustering algorithms.

Clinical data has been employed as the major factor for traditional cancer prognosis. However, this classic approach may be ineffective for analyzing morphologically indistinguishable tumor subtypes. As such, the microarray technology emerges as the promising alternative. Despite a large number of microarray studies, the actual clinical application of gene expression data analysis remains limited due to the complexity of generated data and the noise level. Recently, the integrative cluster analysis of both clinical and gene expression data has shown to be an effective alternative to overcome the above-mentioned problems. This paper presents a novel method for using cluster ensembles that is accurate for analyzing heterogeneous biological data. It overcomes the problem of selecting an appropriate clustering algorithm or parameter setting of any potential candidate, especially with a new set of data. The evaluation on real biological and benchmark datasets suggests that the quality of the proposed model is higher than many state-of-the-art cluster ensemble techniques and standard clustering algorithms. Also, its performance is robust to the parameter perturbation, thus providing a reliable and useful means for data analysts and bioinformaticians. Online supplementary is available at http://users.aber.ac.uk/nii07/bibm2010.

MOTIVATION: It is far from trivial to select the most effective clustering method and its parameterization, for a particular set of gene expression data, because there are a very large number of possibilities. Although many researchers still prefer to use hierarchical clustering in one form or another, this is often sub-optimal. Cluster ensemble research solves this problem by automatically combining multiple data partitions from different clusterings to improve both the robustness and quality of the clustering result. However, many existing ensemble techniques use an association matrix to summarize sample-cluster co-occurrence statistics, and relations within an ensemble are encapsulated only at coarse level, while those existing among clusters are completely neglected. Discovering these missing associations may greatly extend the capability of the ensemble methodology for microarray data clustering. RESULTS: The link-based cluster ensemble (LCE) method, presented here, implements these ideas and demonstrates outstanding performance. Experiment results on real gene expression and synthetic datasets indicate that LCE: (i) usually outperforms the existing cluster ensemble algorithms in individual tests and, overall, is clearly class-leading; (ii) generates excellent, robust performance across different types of data, especially with the presence of noise and imbalanced data clusters; (iii) provides a high-level data matrix that is applicable to many numerical clustering techniques; and (iv) is computationally efficient for large datasets and gene clustering. AVAILABILITY: Online supplementary and implementation are available at: http://users.aber.ac.uk/nii07/bioinformatics2010. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The intuition of data reliability has recently been incorporated into the main stream of research on ordered weighted averaging (OWA) operators. Instead of relying on human-guided variables, the aggregation behavior is determined in accordance with the underlying characteristics of the data being aggregated. Data-oriented operators such as the dependent OWA (DOWA) utilize centralized data structures to generate reliable weights, however. Despite their simplicity, the approach taken by these operators neglects entirely any local data structure that represents a strong agreement or consensus. To address this issue, the cluster-based OWA (Clus-DOWA) operator has been proposed. It employs a cluster-based reliability measure that is effective to differentiate the accountability of different input arguments. Yet, its actual application is constrained by the high computational requirement. This paper presents a more efficient nearest-neighbor-based reliability assessment for which an expensive clustering process is not required. The proposed measure can be perceived as a stress function, from which the OWA weights and associated decision-support explanations can be generated. To illustrate the potential of this measure, it is applied to both the problem of information aggregation for alias detection and the problem of unsupervised feature selection (in which unreliable features are excluded from an actual learning process). Experimental results demonstrate that these techniques usually outperform their conventional state-of-the-art counterparts.

Within the past decades, many fuzzy aggregation techniques, ordered weighted averaging (OWA) in particular, have proven effective for a wide range of information processing tasks, such as decision making, image analysis, database and machine learning. Despite reported successes, their potentials have yet to be explored for the emerging problem of link analysis, which aims to discover similarity and relations amongst objects through their associations. Recently, several link-based similarity methods have been put forward to identifying similar objects in the Internet and publication domains. However, these techniques only take into account the cardinality property of a link structure that is highly sensitive to noise and causes a great number of false positives. In light of such challenge, this paper presents a novel OWA aggregation model that is capable of efficiently deriving a similarity measure through the integration of multiple link properties. The underlying approach is based on the methodology of stress function by which the aggregation behavior can be easily interpreted and modeled. In addition, a semi-supervised method is introduced to assist a user in designing a stress function, i.e. the weighting scheme of link properties, appropriate for a particular link network. The application of the OWA aggregation approach to alias detection is demonstrated and evaluated, against state-of-art link-based techniques, over datasets specifically related to terrorism, publication and email domains.

Combating identity fraud is prominent and urgent since false identity has become the common denominator of all serious crime. Among many identified identity attributes, personal names are commonly falsified or aliased by most criminals and terrorists. Typical approaches to such name disambiguation rely on the text-based similarity measures, which are efficient to some extent, but severely fail to handle highly deceptive and unknown identities. In light of aforementioned shortcoming, this paper presents an intelligent hybrid approach that proficiently combines both content-based and link-based measures of examined names to refine the justification of their similarity. In particular, a new link-based method that exploits multiple link properties is introduced and deployed within the proposed hybrid mechanism. The experimental evaluation of this measure and the hybrid model against other link-based and text-based techniques, over a terrorist-related dataset, significantly indicates their great potentials towards an effective verification system.

Aggregation operators are crucial to integrating diverse decision makerspsila opinion. While minimum and maximum can represent optimistic and pessimistic extremes, an ordered weighted aggregation (OWA) operator is able to reflect varied human attitudes lying between the two using distinct weight vectors. Several weight determination techniques ignore characteristics of data being aggregated. In contrary, data-oriented operators like centered OWA and dependent OWA utilize the centralized data structure to generate reliable weights. Values near the center of a group receive higher weights than those further away. Despite its general applicability, this perspective entirely neglects any local data structures representing strong agreements or consensus. This paper presents a new dependent OWA operator (Clus-DOWA) that applies distributed structure of data or data clusters to determine its weight vector. The reliability of weights created by DOWA and Clus-DOWA operators are experimentally compared in the tasks of classification and unsupervised feature selection.