Search

Although paper shredding is widely used to prevent confidential papers from being misused, still it cannot be considered a convenient process. The size of paper shreds became smaller and smaller, as new methods of shredded paper reassembly and reconstruction are evolving. This paper focuses on clustering, which is a possible phase in the assembly process. This work considers real strip-cut shreds, in addition to images shredded by a simulator in one direction to make strip-cut shreds of different sizes, from wide to narrow shreds, and images shredded in two directions, possibly reflecting cross-cut and micro-cut shreds. K-means is used to cluster shreds, the features tested are gray-level ranges, and the well-known gray-level co-occurrence matrix, invariant moments, segmentation-based fractal texture analysis algorithm, and color moments. The number of shreds grouped in the same cluster with originally adjacent neighbors is used to indicate clustering effectiveness, in addition to the overall accuracy of strip-cut shreds clustering. When the number of clusters is 5, and the k-means experiments run 100 times for 38 images, the overall accuracy of gray-level ranges in simulated strip-cut shreds is 84.87, 89.27, and 93.5 percent in the three different sizes tested, also in cross-cut and micro-cut shreds, gray-level ranges achieve a relatively high number of shreds with 3 and 4 originally adjacent neighbors found in the same cluster. [ABSTRACT FROM AUTHOR]

Clustering is the task of dividing data objects into groups of similar objects. The influence of the specifics of the texts of scientific articles of one subject area for clustering problems has been little studied at present. This article is devoted to the problem of clustering collections of mathematical papers that have different Universal Decimal Classification (UDC) codes. The study was carried out on the collection of mathematical papers published in the journal "Izvestiya VUZov. Matematika" for 10 years. The size of this collection is 623 original papers with different UDC codes. The collection contains subcollections of papers that have the same UDC code. The objective of our research is to analyze the cluster structure of representative subcollections of papers with the same UDC code, which will allow us to evaluate various parameters of the constructed clusters in the future. We have performed the standard pre-processing (tokenization, lemmatization) of Russian math texts. To vectorize text data, we used a method based on - . All vectors were normalized using the L2 norm. To identify optimal hyperparameters and check the quality of clustering, we have used internal efficiency measures such as Silhouette coefficient, Calinski–Harabaz index. The optimal hyperparameters were selected for each method, and then the results of clustering were compared. As a result, we have selected the optimal methods of clustering math papers. [ABSTRACT FROM AUTHOR]

This work presents the results of a transdisciplinary analysis performed on Harward's Almanac (Dublin, 1666), an extremely rare volume currently housed in the National Library of Ireland. The uniqueness and historical value of the Almanac is related to the presence of nineteen handwritten poems, entered by an anonymous scribe. These record textually important English clandestine satire circulating anonymously in Dublin in the late seventeenth and early eighteenth century. Following a comprehensive historical assessment, it appeared evident that the current order of leaves was incorrect. To reconstruct the correct order of the leaves, and hence the likely sequence in which the manuscript poems were inscribed, this study employed a codicological/paleographic analysis complemented by analytical (X-ray fluorescence, XRF) and statistical (Self Organizing Map, SOM) investigation. Specifically, point XRF analysis was carried out for each handwritten page of the Almanac, allowing identification of ink elemental compositions (iron-based ink) and successfully supporting the validity of historical hypotheses on the poems' order of inscription. The statistical organization of XRF data by SOMs allowed easy bi-dimensional visualization of the data set (54 points) and identification of ink similarities, once more validating the historical assessment. [ABSTRACT FROM AUTHOR]

The influence of authors is mostly based on their capacity to form specific self-contained and/or active research communities or topics while also inspiring fruitful spin-off research derived from those communities or topics. Accurately estimating author influence and its indirect effect in inspiring external research creativity must thus be based on very precise community roles as well as topic-role analysis on citing material. In this paper we thus implement, compare, and combine two different kinds of knowledge mapping approaches used to track authors' influence over time—a recent topic-based mapping approach and an original hybrid community-based mapping approach. The experimental context of our study is the analysis of the dynamics and of the influence of the research of Prof. Liu Zeyuan, the most important contributor in the field of Science of Science in China. This analysis focuses on papers citing Professor Liu Zeyuan's research and highlights the full extent of his creative thinking. [ABSTRACT FROM AUTHOR]

A wireless sensor network is a wireless communication network, and each sensor node has several sensors to collect environmental information. Wireless sensor network nodes have limited energy resources and need optimal routing protocols to reduce energy consumption. Failure to reduce energy consumption by sensor nodes reduces network life and efficiency. The main problem in routing is finding optimal paths for sending packets by reducing energy consumption in sensor nodes. This paper proposes an optimal routing method to reduce energy consumption in wireless sensor networks. In the first step, wireless sensor nodes are clustered with the Water strider algorithms (WSA), and cluster heads are selected for routing. In the second step, a mobile sink collects the packets from the cluster heads and sends them to the base station. The mobile sink uses the Ant colony optimization (ACO) algorithms to travel a shorter path between the cluster heads. The authors contribute to presenting a discrete version of the WSA algorithm for cluster head selection to reduce energy consumption. The authors contribute by providing a more comprehensive objective function for clustering network nodes considering error rate, energy consumption, PDR rate, and Euclidean distance. Cluster head traversal with a version of the ACO algorithm to reduce energy consumption and cluster head traversal coding like the TSP problem is the contribution of other authors. The paper aims to reduce energy consumption, reduce the error rate of sending packets and increase the lifetime of the wireless sensor network. Experiments are simulated on several simulated scenarios in Matlab. Criteria such as energy consumption, Packet delivery ratio (PDR), package loss rates, and the number of alive nodes to evaluate the proposed method are used. Experiments show that the proposed algorithm reduces the energy consumption and loss rates of packages of the wireless sensor network by optimally selecting cluster heads and increasing the PDR and number of alive nodes. Comparisons show In terms of energy consumption, Packet delivery ratio (PDR), Loss rates of packages, and the number of alive nodes, the proposed method is more efficient than Particle swarm optimization (PSO), Grey Wolf Optimizer (GWO), Information-centric wireless sensor networks, and Cluster based routing (CBR) routing methods. The PDR index in the proposed method is equal to 97.3% and is higher than PSO, GWO, and CS algorithms. The delay of the proposed method in routing is 25.97%, 5.78%, and 17.98% less than HHO, WOA, and GWO algorithms, respectively. [ABSTRACT FROM AUTHOR]

By the end of 2020, over 16,300 Massive Open Online Courses (MOOCs) from 950 universities worldwide had enrolled over 180 million students. Interest in MOOCs has been matched by significant research on the topic, including a considerable number of reviews. This study uses Machine Learning techniques and human expert supervision to generate a comprehensive systematic literature mapping review that overcomes some limitations of the traditional ones and provides a broader overview of the content and main topics studied in the specialized literature devoted to MOOCs. The sample consisted of 6320 publications automatically classified within six research topics, denominated by human experts: institutional approach, pedagogical approach, evaluation, analytics, participation, and educational resources. The content analysis of the topics identified was conducted using visual network analysis, which supported the identification of different thematic sub-clusters and endorsed the classification. Results from the review show that the lowest production of MOOC papers is within the topics of the pedagogical approach and educational resources. In contrast, participation and evaluation are the most frequent ones. In addition, the most cited papers are on the topics of analytics and resources, being the pedagogical approach and the institutional approach the less cited. This highlights the need for more MOOC research from a pedagogical perspective and calls upon the presence of educators. [ABSTRACT FROM AUTHOR]

Context: Artificial Intelligence (AI) has rapidly developed over the past several decades. Several related AI approaches, such as Machine Learning (ML), have been applied to research on landscape patterns and ecological processes. Objectives: Our goal was to review the methods of AI, particularly ML, used in studies related to landscape ecology and the main topics addressed. We aimed to assess the trend in the number of ML papers and the methods used therein, and provide a synopsis and prospectus of current use and future applications of ML in landscape ecology. Methods: We conducted a systematic literature search and selected 125 papers for review. These were examined and scored according to multiple criteria regarding methods and topic. We applied quantitative statistical methods, including cluster analysis based on titles, abstracts, and keywords and a non-metric multidimensional scaling based on attributes assigned during the review. We used Random Forests machine learning to describe the differences between identified clusters in terms of the topics and methods they included. Results: The most frequent method found was Random Forests, but it is noteworthy to mention the increasing popularity of tools related to Deep Learning. The topics cover both ecologically oriented issues and the landscape-human interface. There has been a rapid increase in ML and AI methods in landscape ecology research, with Deep Learning and complex multi-step pipeline AI methods emerging in the last several years. Conclusions: The rapid increase in the number of ML papers in landscape ecology research, and the range of methods employed in them, suggest explosive growth in application of these methods in landscape ecology. The increase of Deep Learning approaches in the most recent years suggest a major change in analytical paradigms and methodologies that we feel may transform the field and enable analyses of more complex pattern process relationships across vaster data sets than has been possible previously. [ABSTRACT FROM AUTHOR]

An efficient variant of the recent sea horse optimizer (SHO) called SHO-OBL is presented, which incorporates the opposition-based learning (OBL) approach into the predation behavior of SHO and uses the greedy selection (GS) technique at the end of each optimization cycle. This enhancement was created to avoid being trapped by local optima and to improve the quality and variety of solutions obtained. However, the SHO can occasionally be vulnerable to stagnation in local optima, which is a problem of concern given the low diversity of sea horses. In this paper, an SHO-OBL is suggested for the tackling of genuine and global optimization systems. To investigate the validity of the suggested SHO-OBL, it is compared with nine robust optimizers, including differential evolution (DE), grey wolf optimizer (GWO), moth-flame optimization algorithm (MFO), sine cosine algorithm (SCA), fitness dependent optimizer (FDO), Harris hawks optimization (HHO), chimp optimization algorithm (ChOA), Fox optimizer (FOX), and the basic SHO in ten unconstrained test routines belonging to the IEEE congress on evolutionary computation 2020 (CEC'20). Furthermore, three different design engineering issues, including the welded beam, the tension/compression spring, and the pressure vessel, are solved using the proposed SHO-OBL to test its applicability. In addition, one of the most successful approaches to data transmission in a wireless sensor network that uses little energy is clustering. In this paper, SHO-OBL is suggested to assist in the process of choosing the optimal power-aware cluster heads based on a predefined objective function that takes into account the residual power of the node, as well as the sum of the powers of surrounding nodes. Similarly, the performance of SHO-OBL is compared to that of its competitors. Thorough simulations demonstrate that the suggested SHO-OBL algorithm outperforms in terms of residual power, network lifespan, and extended stability duration. [ABSTRACT FROM AUTHOR]

Topic models are a well known clustering approach for textual data, which provides promising applications in the bibliometric context for the purpose of discovering scientific topics and trends in a corpus of scientific publications. However, topic models per se provide poorly descriptive metadata featuring the discovered clusters of publications and they are not related to the other important metadata usually available with publications, such as authors affiliation, publication venue, and publication year. In this paper, we propose a methodological approach to topic modeling and post-processing of topic models results to the end of describing in depth a field of research over time. In particular, we work on a selection of publications from the international statistical literature, we propose an approach that allows us to identify sophisticated topic descriptors, and we analyze the links between topics and their temporal evolution. [ABSTRACT FROM AUTHOR]

A dataset containing 111,616 documents in astronomy and astrophysics (Astro-set) has been created and is being partitioned by several research groups using different algorithms. For this paper, rather than partitioning the dataset directly, we locate the data in a previously created model of the full Scopus database. This allows comparisons between using local and global data for community detection, which is done in an accompanying paper. We can begin to answer the question of the extent to which the rest of a large database (a global solution) affects the partitioning of a smaller journal-based set of documents (a local solution). We find that the Astro-set, while spread across hundreds of partitions in the Scopus map, is concentrated in only a few regions of the map. From this perspective there seems to be some correspondence between local information and the global cluster solution. However, we also show that the within-Astro-set links are only one-third of the total links that are available to these papers in the full Scopus database. The non-Astro-set links are significant in two ways: (1) in areas where the Astro-set papers are concentrated, related papers from non-astronomy journals are included in clusters with the Astro-set papers, and (2) Astro-set papers that have a very low fraction of within-set links tend to end up in clusters that are not astronomy-based. Overall, this work highlights limitations of the use of journal-based document sets to identify the structure of scientific fields. [ABSTRACT FROM AUTHOR]

A network of wireless sensors (WSN) is an outstanding technology that can aid in the various applications. Batteries run the sensor nodes those are used in WSN. The battery is impossible to charge or repair, so the most valuable resource for wireless sensor networks is power. Over the years, several strategies have been invented and used to preserve this precious WSN resource. One of the most successful approach for this purpose has turned out to be clustering. The aim of this paper is to suggest an effective technique for choosing cluster heads in WSNs to increase the lifetime of the network. To accomplish this task, Grey Wolf Optimizer (GWO) technique has been used. The general GWO was updated in this paper to meet the particular purpose of cluster head selection in WSNs. In this article, we have considered eleven attributes in the fitness function for the proposed algorithm. The simulation is carried out under different conditions. The results obtained show that the proposed protocol is superior in terms of energy consumption and network lifetime by evaluating the proposed protocol (i.e. CH-GWO protocol) with some well-existing cluster protocols. The suggested protocol forms energy-efficient and scalable clusters. [ABSTRACT FROM AUTHOR]

Clustering is a very important method to analyze HIN. Thus, several HIN clustering algorithms have been proposed and all these algorithms are based on meta-paths. Meta-path can be used to describe the relationship of target objects. Even though the relationship of target objects are fully considered by these meta-path based algorithms, the directly connected neighbors of target objects are neglected by them. These neglected directly connected neighbors are not target objects, but they contain plenty of useful information for finding clusters of target objects. So, while performing clustering based on HIN, these neglected neighbors should be considered. To achieve the goal, in this paper, a new HIN clustering algorithm is proposed. The proposed algorithm tries to build a vector for each target object. The clustering task is fulfilled based on these vectors. During the vector building process, the neighbors of all the target objects are considered. As clustering result of HIN is affected by different factors, such as neighbors of target objects and different kinds of meta-paths, several similarity matrices are built in the proposed algorithm. Each matrix is corresponding to a specific factor. Besides, every matrix will be assigned a weight value. These weight values are used to represent the relative importance of factors. At the same time, in the proposed algorithm, a new training method is adopted to calculate the vectors and the weight values. [ABSTRACT FROM AUTHOR]

In this paper, preliminary studies on the failure analysis of hybrid composite materials utilizing acoustic emission and machine learning are presented. The main purpose of this study was to analyze the possibilities of using machine learning techniques as a way to better cluster the data obtained from acoustic emission. In this paper, we focus on data preparation, feature extraction (Laplacian score), determination of cluster number (Caliński–Harabasz, Silhouette, and Davies–Bouldin), and testing three clustering techniques, namely K-means, fuzzy C-means, and spectral clustering. The dataset was obtained by testing fiber metal laminates—composites consisting of metal and composite layers. Two experimental tests were realized on pre-cracked rectangular specimens—one with loading in mode I and one with loading in mode II (DCB—double cantilever beam and ENF—end-notch flexural test). Elastic waves were recorded during these tests via an acoustic emission system. Preliminary studies show that the proposed method can be used successfully to cluster data obtained in this way. The obtained dataset was split into 3 clusters (for the ENF test) and 5 clusters (DCB test). In the next stages of the research campaign, based on the presented results, we intend to change the approach to semi-supervised by running additional single-cause damage tests to enhance the achieved results and enable easier damage recognition. [ABSTRACT FROM AUTHOR]

Fuzzy clustering has been useful in capturing the uncertainty present in the data during clustering. Most of the c-Means algorithms such as FCM (Fuzzy c-Means), IFCM (Intuitionistic Fuzzy c-Means), and the recently reported PIFCM (Probabilistic Intuitionistic Fuzzy c-means) randomly initialize cluster centroids. Performance of these techniques is very reliant on the initialized cluster centroids. So, a good initialization technique can significantly affect the cluster formation. Recently, density-based initialization technique for FCM (DFCM) was proposed, which initializes datapoints with high density as cluster centroids. In DFCM, points within some distance contribute in the density of the data points. In this paper, we propose a new way to compute fuzzy density of datapoints based on the distance measure. Uncertainty can be better captured by intuitionistic fuzzy set (IFS) and interval-valued IFS (IVIFS). Thus, we propose a new density-based initialization technique for IFCM, called 'Density based Intuitionistic Fuzzy c-Means (DIFCM) Algorithm'. The proposed DIFCM has been further developed for IVIFS, which we term 'Interval-valued Density based Intuitionistic Fuzzy c-Means (IVDIFCM) Algorithm', is also introduced in this paper. PIFCM incorporates probabilistic weights between membership, non-membership and hesitancy component. In this paper, we also introduce the density based initialized cluster centroids for PIFCM algorithm to propose the 'Density Based Probabilistic Intuitionistic Fuzzy c-Means (DPIFCM) Algorithm'. There were many clustering approaches based on IFSs but there do not exist any literature review on the IFS based clustering approaches. Therefore, this article also provides a detailed review of the recently proposed clustering algorithms based on IFS theory. Experiments over various UCI datasets proves that our proposed algorithms has better clustering results than their existing counterparts. [ABSTRACT FROM AUTHOR]

Clustering validity indices are designed for evaluating the performance of clustering algorithms in terms of quality of clusters. They are also used for detecting the correct number of clusters. Numerous clustering validity indices have been proposed in the literature for different types of data. However, to the best of our knowledge, there are a few well-known clustering validity indices for graph data. In this paper, we propose a new clustering validity index for graph data called the node-based clustering validity index. The main characteristic of our proposed index is that it captures the exclusive contribution of each node to the separation and compactness of the clusters in a graph. This characteristic gives an advantage to the method which is the detection of the correct number of clusters for cases that two sets of nodes are bonded to each other with a single node. We provide an illustrative example to show that many existing validity indices fail to detect the correct number of clusters in such cases. We evaluate the performance of our proposed index with several experiments on different real-world and synthetic graphs. Experiments show that our proposed node-based index outperforms the existing validity indices for different graph data. [ABSTRACT FROM AUTHOR]

This paper presents a novel clustering approach that utilizes variational autoencoders (VAEs) with disentangled representations, enhancing the efficiency and effectiveness of clustering. Traditional VAE-based clustering models often conflate generative and clustering information, leading to suboptimal clustering performance. To overcome this, our model distinctly separates latent representations into two modules: one for clustering and another for generation. This separation significantly improves clustering performance. Additionally, we employ augmented data to maximize mutual information between cluster assignment variables and the optimized latent variables. This strategy not only enhances clustering effectiveness but also allows the construction of latent variables that synergistically combine clustering information from original data with generative information from augmented data. Through extensive experiments, our model demonstrates superior clustering performance without the need for pre-training, outperforming existing deep generative clustering models. Moreover, it achieves state-of-the-art clustering accuracy on certain datasets, surpassing models that require pre-training. [ABSTRACT FROM AUTHOR]

With the increasing deployment of FANETs, efficient routing algorithms play a crucial role in ensuring reliable and optimal communication among UAVs. This paper presents a novel Clustered Learning Automata-based Routing Algorithm (called CLARA) for FANETs. The proposed algorithm is designed to enhance the performance of routing by leveraging the capabilities of learning automata. The CLARA consists of five key phases. In the first phase, actions, states, and a Q-Table are initialized to facilitate the learning process. Subsequently, in the second phase, drones within the FANET are clustered based on their geographic location and connection characteristics. This clustering enables efficient management of UAVs and improves network performance. In the third phase, the algorithm selects the next hop for data packet transmission, considering factors such as connectivity and reliability. This selection process ensures robust and efficient data delivery within the network. The fourth phase focuses on Q-Table updating and energy management, which optimizes resource allocation and prolongs network lifetime. Finally, in the fifth phase, the cluster head is selected based on its remaining energy, ensuring effective leadership within each cluster. This dynamic selection process enables the efficient distribution of responsibilities and enhances network stability. Simulation results demonstrate the superiority of the CLARA approach compared to existing methods, including OLSR, AODV, and Q-FANET. CLARA algorithm shows its superiority in the criteria of control overhead, routing overhead, computational overhead of energy consumption, network lifetime, PDR, and end-to-end delay. [ABSTRACT FROM AUTHOR]

The k-means algorithm is a prevalent clustering method due to its simplicity, effectiveness, and speed. However, its main disadvantage is its high sensitivity to the initial positions of the cluster centers. The global k-means is a deterministic algorithm proposed to tackle the random initialization problem of k-means but its well-known that requires high computational cost. It partitions the data to K clusters by solving all k-means sub-problems incrementally for all k = 1 , ... , K . For each k cluster problem, the method executes the k-means algorithm N times, where N is the number of datapoints. In this paper, we propose the globalk-means++ clustering algorithm, which is an effective way of acquiring quality clustering solutions akin to those of global k-means with a reduced computational load. This is achieved by exploiting the center selection probability that is effectively used in the k-means++ algorithm. The proposed method has been tested and compared in various benchmark datasets yielding very satisfactory results in terms of clustering quality and execution speed. [ABSTRACT FROM AUTHOR]

The residualization of public rented housing is a prevalent phenomenon throughout Europe, and strongly present in the small and strongly means-tested social housing sector in Norway. In this article, we discuss the contested geographical dimension of residualization. Scientific studies of the geographical and locational aspects of social housing are scare in Norway and modest internationally. Based on qualitative interviews with representatives of social housing administrators in the fifteen largest urban municipalities in Norway, this paper contributes to the literature by exploring how these social housing bureaucrats perceive, reflect on, and respond to, questions related to the spatial localization of residual social housing. Does it matter where social housing is located? What are the consequences of the geography of social housing for tenants, their neighbours, and the wider socio-spatial development of cities? These are questions pondered in the interviews. In our qualitative analysis, we identify three broad themes. First, the theme of the internal social milieu – inclusive communities versus neighbour complaints and conflicts in the public housing projects. Second, the theme of neighbourhood effects; how concentrated poverty is influencing the local community in general and the upbringing of children in particular. Third, the theme of response from external neighbours and communities, in the form of either predominantly exclusive strategies (NIMBYism – Not in My Backyard), but also less prevalent inclusive strategies like (PHIMBYism – Public Housing In My Backyard). [ABSTRACT FROM AUTHOR]

As an extension of Fuzzy C-Means (FCM), Evidence C-Means (ECM) is proposed in the framework of Dempster–Shafer theory (DST) and has been applied to many fields. However, the objective function of ECM involves only the distortion between the object and the prototype, which relies heavily on the initial prototype. Therefore, ECM may encounter the problem of local optimization. To solve this problem, this paper introduces ECM with Particle Swarm Optimization (PSO) initialization to determine the initial clustering centroids, and proposes Particle Swarm Optimization-based Evidential C-Means (PSO-ECM), which reduces the influence of bad initial prototypes and improves the local optimality problem of ECM. PSO-ECM is compared with three other clustering algorithms in four experiments and with ECM on a noise-containing dataset. According to the experimental results, PSO-ECM performs well in terms of different clustering validity metrics compared with existing clustering algorithms, has high stability of clustering, and can effectively and stably cluster noise-containing datasets and accurately identify outlier points. [ABSTRACT FROM AUTHOR]

Money laundering identification (MLI) is a challenging task for financial AI research and application due to its massive transaction volume, label sparseness, and label bias. Most of the existing MLI methods focus on individual-level abnormal behavior while neglecting the community factor that money laundering is a collaborative group crime. Furthermore, the massive volume of transactions and the issue of label shifting also impede the application of supervised or semi-supervised models. To this end, this paper proposes an efficient community-oriented algorithm, namely SEGE, to identify money laundering based on structural entropy minimization (SEM) with graph embedding in an unsupervised approach. Experiments on both a private real-world money laundering network and a public synthetic dataset show that our SEGE algorithm derives prominent performance and outperforms the parameterized learning-based graph representation methods. Moreover, we find that there are pervasive sub-communities in the real-world money laundering network. Based on our local algorithm, we propose a real combat strategy against the money laundering group, in which when we have several scattered suspicious accounts in the transaction network, we are able to retrieve the whole money laundering group by the union of sub-communities with both high precision and high recall rates. [ABSTRACT FROM AUTHOR]

Although artificial intelligence has become part of everyone's real life, a trust crisis against such systems is occurring, thus increasing the need to explain black-box predictions, especially in the military, medical, and financial domains. Modern eXplainable Artificial Intelligence (XAI) techniques focus on benchmark datasets, but the cognitive applicability of such solutions under big data settings is still unclear due to memory or computation constraints. In this paper, we extend a model-agnostic XAI methodology, named Cluster-Aided Space Transformation for Local Explanation (CASTLE), to be able to deal with high-volume datasets. CASTLE aims to explain the black-box behavior of predictive models by combining both local (i.e., based on the input sample) and global (i.e., based on the whole scope for action of the model) information. In particular, the local explanation provides a rule-based explanation for the prediction of a target instance as well as the directions to update the likelihood of the predicted class. Our extension leverages modern big data technologies (e.g., Apache Spark) to handle the high volume, variety, and velocity of huge datasets. We have evaluated the framework on five datasets, in terms of temporal efficiency, explanation quality, and model significance. Our results indicate that the proposed approach retains the high-quality explanations associated with CASTLE while efficiently handling large datasets. Importantly, it exhibits a sub-linear, rather than exponential, dependence on dataset size, making it a scalable solution for massive datasets or in any big data scenario. [ABSTRACT FROM AUTHOR]

Face clustering has applications in organizing personal photo album, video understanding and automatic labeling of data for semi-supervised learning. Many existing methods cannot cluster millions of faces. They are either too slow, inaccurate, or need a lot memory. In our paper, we proposed a two stage unsupervised clustering algorithm which can cluster millions of faces in minutes. A rough clustering using greedy Transitive Closure (TC) algorithm to separate the easy to locate clusters, then a more precise non-greedy clustering algorithm is used to split the clusters into smaller clusters. We also developed a set of omni-supervised transformations that can produce multiple embeddings using a single trained model as if there are multiple models trained. These embeddings are combined using simple averaging and normalization. We carried out extensive experiments with multiple datasets of different sizes comparing with existing state of the art clustering algorithms to show that our clustering algorithm is robust to differences between datasets, efficient and outperforms existing methods. We also carried out further analysis on number of singleton clusters and variations of our model using different non-greedy clustering algorithms. We did trained our semi-supervised model using the cluster labels and shown that our clustering algorithm is effective for semi-supervised learning. [ABSTRACT FROM AUTHOR]

Belief and plausibility functions based on evidence theory (ET) have been widely used in managing uncertainty. Various generalizations of ET to fuzzy sets (FSs) have been reported in the literature, but no generalization of ET to q-rung orthopair fuzzy sets (q-ROFSs) has been made yet. Therefore, this paper proposes a novel, simple, and intuitive approach to distance and similarity measures for q-ROFSs based on belief and plausibility functions within the framework of ET. This research addresses a significant research gap by introducing a comprehensive framework for handling uncertainty in q-ROFSs using ET. Furthermore, it acknowledges the limitations inherent in the current state of research, notably the absence of generalizations of ET to q-ROFSs and the challenges in extending belief and plausibility measures to certain aggregation operators and other generalizations including Hesitant fuzzy sets, Bipolar fuzzy sets, Fuzzy soft sets etc. Our contribution lies in the proposal of a novel approach to distance and similarity measures for q-ROFSs under ET, utilizing Orthopairian belief and plausibility intervals (OBPIs). We establish new similarity measures within the generalized ET framework and demonstrate the reasonability of our method through useful numerical examples. Additionally, we construct Orthopairian belief and plausibility GRA (OBP-GRA) for managing daily life complex issues, particularly in multicriteria decision-making scenarios. Numerical simulations and results confirm the usability and practical applicability of our proposed method in the framework of ET. [ABSTRACT FROM AUTHOR]

The purpose of clustering is to partition data similar with each other into a same group and partition data dissimilar with each other into different groups. However, in most existing fuzzy clustering approaches, the membership degrees of an individual belonging to different clusters are relied on the different distances between the individual and different cluster centroids, the similarity between the individual and data in different clusters are ignored, besides, the outliers cannot be distinguished effectively. For improving the clustering performance and addressing the problems in most existing fuzzy clustering approaches, based on the concept that data close to each other should be grouped together and be assigned to a same cluster, in this paper, we present a novel efficient fuzzy clustering approach, in which the adaptive local neighborhood information of each data referring to different clusters is taken into consideration; an entirely new idea that the membership degree values of an individual referring to different clusters should not only depend on the distances between the individual and different cluster centers, but also rely on the distances between the individual and several nearest neighbor data in different clusters is put forward; a new scheme to search for the outliers is presented, a method for identifying the different importance of different features is introduced. Experiments on synthetic and publicly real-world datasets were conducted, the clustering results show that the approach put forward can yield consistently satisfying clustering performance and has significant advantages. [ABSTRACT FROM AUTHOR]

Given the rise in demand for cloud computing in the modern era, the effectiveness of resource utilization is eminent to decrease energy footprint and achieve economic services. With the emerging machine learning and artificial intelligence techniques to model and predict, it is essential to explore a principal method that provides the best solution for the accurate provisioning of forthcoming requests in a cloud data center. Recent studies used machine learning and other advanced analytics to predict resource usage; however, these do not consider long-range dependencies in the time series, which is essential to capture for better prediction. Further, they show limitations in handling noise, missing values, and outliers in datasets. In this paper, we explored the problem by studying three techniques that enabled us to answer improvements in short-term forecasting of physical machines' resource usage if the above factors are considered. We evaluated the predictions using Transformer and Informer deep learning models that cover the above aspects and compared them with the Long short-term memory (LSTM) model. We used a real-world Google cluster trace usage dataset and employed Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) algorithm to select heterogeneous machines. The evaluation of the three models depicts that the Transformer architecture that considers long-range dependencies in time series and shortcomings with datasets shows improvement in forecasting with 14.2% reduction in RMSE than LSTM. However, LSTM shows better results for some machines than the Transformer, which depicts the importance of input sequence order. The Informer model, which considers both dependencies and is a hybrid of LSTM and Transformer, outperformed both models with 21.7% from LSTM and 20.8% from Transformer reduction in RMSE. The results also depict Informer model consistently performs better than the other models across all subsets of the dataset. Our study proves that considering long-range dependencies and sequence ordering for resource usage time series improves the prediction. [ABSTRACT FROM AUTHOR]

Among the declining industries, for example music industry, some have been revived by information technology (IT). At the same time, in academic fields, some have expected co-evolutions between IT and other fields to cause the resurgence of either field. In this research, the clustering of citation networks with 14,438 academic papers resulted in the identification of 28 academic fields in the areas "Computer Science" or "Information Science and Library Science." Co-evolutions between these 28 fields and citing fields to the 28 fields were evaluated by an investigation of contents; a methodology to search co-evolutions was also proposed. This paper proposes that pairs of academic fields (with both high correlation and high dissimilarity) co-evolve, and some co-evolving pairs of academic fields were found. This research contributes to the discovery of the co-evolution between academic fields. [ABSTRACT FROM AUTHOR]

Spatio-temporal (S-T) analysis is not typical in structural monitoring applications of buildings and infrastructure. However, monitoring always includes the temporal component, and observations are often captured in specific locations. In other words, a monitoring dataset could also be considered a spatio-temporal archive, notwithstanding that not all monitoring applications can benefit from S-T processing methods. The paper discusses spatio-temporal analysis using the structural monitoring dataset of the Cathedral of Milan, which has an archive of vertical settlements collected from more than 50 years of measurements. The proposed methods can be adapted and extended for other structural monitoring applications, including single buildings, infrastructure, and the environmental level. The cases of pure temporal (T) and spatial (S) analyses are also discussed, comparing the different approaches, illustrating the pros and cons, and describing the opportunities of the S-T combined workflow. The paper specifically focuses on different typologies of S-T processing: data visualization and exploration techniques, clustering, change detection, prediction, and forecasting. The proposed algorithms were all implemented within the R open-source programming language. They can be replicated (and adapted) for other structural monitoring datasets featuring spatio-temporal correlation. [ABSTRACT FROM AUTHOR]

Context: The recent research trend has highlighted that multiple stakeholders are involved during requirement gathering in agile software development. Hence, leading to an increased number of duplicate user stories in agile product backlog during requirement gathering. Objective: The objective of this paper is to evaluate the existing techniques employed in identifying and eliminating the duplicate user stories from agile product backlog and to overcome the existing gaps with the help of a newly proposed clustering algorithm. Method: An agile user story is expressed as a function of input and output parameters. That said multiple user stories having similar set of input parameters are most likely to be duplicate causing a redundancy. The newly proposed algorithm is used for clustering user stories having similar set of input parameters through various iterations and then removing the identified duplicate user stories from agile product backlog. This paper also introduces the concept of mass clustering which means clustering a number of user stories in single run. Results: Experimental results prove the proposed model is capable of handling small and large releases ranging between 100 to 1000 user stories with similar efficiency. The proposed clustering algorithm outperformed the clustering algorithms and resulted in 37% decrease in agile product backlog by eliminating duplicate user stories causing redundancy. The experimental results are obtained from the logs of the MATLAB tool. However, the provided algorithm is generic in nature and can be implemented using R, Python or SAS programming tools. The provided algorithms employs proven matrix operations. Conclusion: The proposed clustering algorithm overcomes the limitation of existing user story management methods and clearly out performs when compared with other clustering algorithms. Finally, this paper gives recommendations about the usage of the provided clustering algorithm during agile release planning for eliminating duplicate user stories from agile product backlog. [ABSTRACT FROM AUTHOR]

In wireless sensor networks (WSNs), sensor nodes were considered to be an integral part of IoT (Internet of Things) for sensing and monitoring the environment. The IoT-based applications need to be optimized regarding the changing requirements of users as everything is connected via the internet. In today's era, where every day new technologies were rebuilt where sensor nodes plays an important role on it. In every field, whether it is healthcare, smart agriculture, smart home appliances, smart traffic, or smart city sensors were deployed for sensing their environment, collecting data from them, and forwarding it to the servers. These sensor nodes were made up of non-rechargeable power batteries, as a fact efficient energy consumption of these batteries becomes vital. In WSN, efficient energy consumption is still an issue, and its solutions were given by many researchers among them, clustering is considered to be more effective in this domain. For efficient energy consumption, multi-attributes of cluster head selection need to be considered and proper coordination among the conflicting nature of multi-attributes needs to be done. In this paper, we have proposed PROMETHEE II and modified AHP together for cluster heads selection by considering multi-attributes. Twenty-one attributes were considered including connectivity, distance to the base station, residual energy, member nodes, and many more. Being conflicting in nature, proper coordination among these attributes has been done and optimal cluster heads were selected modified for data transmissions. In this paper, modified AHP has been compared with our proposed modified PROMETHEE II and AHP for understanding the significance of this integration. Results is evaluated in terms of energy consumption, network lifetime, and load balancing and it also validate that our proposed approach outperforms with modified AHP and other existing algorithms. Our proposed algorithm enriches network lifetime by balancing the load among sensor nodes which leads to efficient energy consumption. [ABSTRACT FROM AUTHOR]

This paper proposes a parsimonious model of network formation with introductions in the presence of intermediation rents. Introductions allow two nodes to form a new connection on favorable terms with the help of a common neighbor. The decision to form links via introductions is subject to a trade-off between the gains from having a direct connection at lower cost and the potential losses for the introducer from lower intermediation rents. When nodes take advantage of introductions, stable networks tend to exhibit a minimum amount of clustering. At the same time, intermediary nodes have incentives to protect their position, and stable networks can exhibit nodes exploiting structural holes, that is, bridges across otherwise unconnected parts of the network earning intermediation rents. [ABSTRACT FROM AUTHOR]

IoT presents a new technology that aims to interconnect numerous smart devices to provide operators with many smart services. Generally, wireless sensor networks (WSNs) are the key paradigm that ensures Internet of Things (IoT) applications. Due to the limitation of energy sensor resources, the IoT-based WSN is an energy efficiency strategy. In this context, many strategies for energy conservation have been proposed in the literature. Due to the absence of a comparative study of these methods in the literature, the present paper proposes to study 60 recent energy-efficient methods dedicated to IoT applications based on WSNs. Paper selection is accomplished based on the best citation through the Web of Science, Scopus, and SCI indexed databases. First, the different approaches proposed in the literature are classified into three categories: routing methods, clustering methods, and miscellaneous methods. Routing methods are classified into seven types: multipath routing, distributed routing, centralized routing, survivable path routing, opportunistic routing, SDN- based routing, and QoS-based routing. Moreover, clustering approaches are organized into four types: fuzzy clustering, random clustering, weighted clustering, and hybrid clustering. For the Miscellaneous part, three phases are discussed: the physical layer, MAC layer, and security. Furthermore, a comparative study is performed concerning types, performance metrics, applications, simulators, and future research directions. This analysis helps field researchers to learn the latest methods in this field. Established methods or testing those inside a testbed may be an essential path for potential studies. [ABSTRACT FROM AUTHOR]

Clustering is a technique that segregates a provided dataset into homogenous groups in accordance with the provided features. It aims to determine a structure in a group of unlabelled data. Cluster analysis is an unsupervised learning technology that determines the interesting patterns in data objects without class labels. K mode clustering algorithm seems to be effective in clustering categorical data due to its easy implementation and capability to handle the massive amount of data. But because of its random selectivity of initial centroids, it gives the local optimum solution. The main contribution of the paper is to evaluate the performance of clustering on the various dataset with the proposed system. The proposed method utilizes a genetic-based Metaheuristic encircle algorithm to select enriched features and novel dynamic K modes clustering based on Dimensionality Reduced PSO for clustering process with better computational time. The encircling Prey concept has been incorporated to choose the fitness function and overcome the genetic algorithm limitations in feature selection. This paper integrated the k-modes algorithm with particle swarm optimization algorithm to obtain a global optimum solution and update the initial centroid. Several dataset utilized for the evaluation of the proposed work has been found to achieve low accuracy in the previous work. But the proposed approach's effectiveness has been proved to be better by performing a comparative analysis with the state of art methods in terms of performance metrics such as F1 score, accuracy, NMI. [ABSTRACT FROM AUTHOR]

Recently, wireless sensor networks (WSNs) are becoming very famous as they are inexpensive and easy to maintain and manage. The network contains a group of sensor nodes, which are capable of sensing, computing, and transmitting. Energy efficiency is one of the most important challenging problems in WSN. Sensor nodes have inadequate energy and installed in remote areas. Hence, it is difficult to restore the batteries in WSN. Therefore, to maximize the network lifetime, appropriate clustering techniques and cluster head (CH) selection methods should be implemented. The main idea behind the clustering technique is that it clusters the sensor nodes and reduces the composed data simultaneously and then, it broadcasts the data. In this process, CH selection is an essential part. Therefore, this survey paper provides an overview of the clustering techniques for reducing energy consumption by reviewing several CH selection techniques in WSN that provide high energy efficiency. Several techniques have been employed for CH selection based on partitional clustering, optimization, low-energy adaptive clustering hierarchy, hierarchical, distributed, and other classification methods. Finally, an analysis is done based on the implementation tools, metrics employed, accuracy, and achievements of the considered CH selection techniques. [ABSTRACT FROM AUTHOR]

The intrusion detection system is responsible for revealing different intrusion activities, including the denial of service, man-in-middle, Mirai, Scan, and other types of intrusion activities. It is used in many applications, including the smart home Internet of Things networks, where security risks threaten the privacy of individuals. In this context, many works were proposed for detecting and classifying the different types of attacks. However, many challenges are identified for this type of problem, such as the large amount of data available, the imbalanced nature of the data, and the quality of detection and classification outcomes. This paper aims to address these challenges by proposing an approach that considers a metaheuristic-based random weight neural network to detect intrusion activities and classify the different types and subtypes of activities. The following points summarize the contribution of this paper. First, the automatic tuning of the neural network parameters where the weights, biases, regularization value, the number of neurons, and the type of activation function are optimized by different metaheuristic algorithms to produce high-quality results. Second, the proposed approach adopts a clustering with reduction technique to tackle the challenge of processing large volumes of data. Third, oversampling the dataset is also embedded in the proposed approach to avoid a biased classification of the majority class. The experiments are conducted based on a large dataset with more than half a million instances. The results show that the proposed approach outperforms the other classification approaches in geometric mean (G-Mean) and has promising results. [ABSTRACT FROM AUTHOR]

Wireless Sensor Networks (WSN) are composed of small sensor nodes that either transmit their sensed data to the sink node directly or transmit it to its respective cluster head, which then transmits it to the sink node. However, this consumes a lot of network bandwidth and energy from the constrained sensor nodes. To address these constraints, Mobile Agents (MA) paradigm can be used in WSNs, which may lead to better energy and bandwidth conservation. When a single mobile agent is insufficient to complete a task, multiple mobile agents can be deployed to perform in parallel and reduce network latency. The set of sensor nodes and their sequence that MAs must migrate to complete a task is called an itinerary. The planning of the itinerary is the most prominent and significant issue related to the MA-based system, including the determination of an appropriate number of MAs to be dispatched, determining the set of sensor nodes and their sequence to be visited by MAs. This paper proposes a fuzzy-based algorithm to partition Wireless Sensor Networks into a set of sensor nodes, called domains, for enhancing the efficiency of the WSN in terms of its prolonged operation. Experimental evaluations are conducted to compare the proposed algorithm with benchmarked algorithms. The paper suggests that the proposed algorithm's integration with MA-based systems can enhance their performance and prolong the WSN's lifetime. [ABSTRACT FROM AUTHOR]

With the advancement of technology such as video monitoring, Internet-of-things, cloud, and machine learning, Industry 4.0 is working continuously to ensure the security of workers. The workers are equipped with sensors to analyze their activities. In general, the recognition of human activities in cloud-based industrial scenario is leveraged to monitor the safety of the workers. This paper introduced a new optimal clustering method for the activity recognition of workers in industry using cloud based IoT environment. The proposed method uses the temporal and spatial features of human workers in industry. The proposed method is tested on publicly available dataset of different activities maintained into three groups, namely movement, gestures, and object handling, in the context of the medium and small industrial environment. The experimental findings validate that the proposed method achieves 80.2 % , 81.05 % and 80.19 % of average accuracy for movement, gesture, and object handling activities, which clearly outperformed the fuzzy c-means, particle-swarm optimization, and HMM-based activity recognition methods. [ABSTRACT FROM AUTHOR]

This paper presents an automated approach for providing ranked lists of outliers in observed demand to support analysts in network revenue management. Such network revenue management, e.g. for railway itineraries, needs accurate demand forecasts. However, demand outliers across or in parts of a network complicate accurate demand forecasting, and the network structure makes such demand outliers hard to detect. We propose a two-step approach combining clustering with functional outlier detection to identify outlying demand from network bookings observed on the leg level. The first step clusters legs to appropriately partition and pools booking patterns. The second step identifies outliers within each cluster and uses a novel aggregation method across legs to create a ranked alert list of affected instances. Our method outperforms analyses that consider leg data without regard for network implications and offers a computationally efficient alternative to storing and analysing all data on the itinerary level, especially in highly-connected networks where most customers book multi-leg products. A simulation study demonstrates the robustness of the approach and quantifies the potential revenue benefits from adjusting demand forecasts for offer optimisation. Finally, we illustrate the applicability based on empirical data obtained from Deutsche Bahn. [ABSTRACT FROM AUTHOR]

The ever-increasing number of cloud services has led to the service's identification problem. It has become difficult to provide users with cloud services that meet their functional and non-functional requirements, especially as many cloud services offer the same or similar functionality but with different execution constraints (cloud characteristics, QoS, price, and so on). Service recommendation systems can solve the service's identification problem by helping users to retrieve the right cloud services according to their desired needs. However, the majority of service recommendation systems rely on user feedback to locate the user's neighbors, predict missing ratings, and rank the recommended services. As a result, users' rating histories might cause three major problems: cold start, data sparsity, and malicious attack. In order to deal with these issues, we propose in this paper a hybrid recommendation approach, called "HRPCS", that provides a list of personalized cloud services to the active user. This approach is based on user and service clustering. In this approach, cloud services are recommended based on the user's needs (functional and non-functional) and QoS preferences. Then, the services are ranked according to their prices and credibility. Further, the proposed approach returns a list of diversified cloud services. The experimental results confirmed our expectations and proved the effectiveness of our approach. [ABSTRACT FROM AUTHOR]

This paper presents a new cluster-based quality of service routing protocol (C-QoSRP) designed for VANET in a highway setting, offering the benefit of wide-range communication facilitated by road-side units (RSUs). The protocol incorporates the K-nearest neighbor (KNN) algorithm to optimize cluster stability by combining mobility and link quality metrics. A noteworthy aspect of this approach is that in the event of a cluster head (CH) failure, C-QoSRP employs RSUs as temporary cluster heads (TCHs). The TCH assumes the CH's responsibilities to ensure reliable communication within the cluster. Unlike traditional protocols that primarily focus on CH stability, C-QoSRP aims to minimize transmission delays and improve overall network stability. Simulation results demonstrate that C-QoSRP protocol surpasses existing ones in terms of QoS parameters such as packet delivery ratio, end-to-end delay, throughput, overhead and Cluster stability. [ABSTRACT FROM AUTHOR]

COVID-19 has triggered fundamental shifts in how people live, work and consume. The crisis stimulated varied responses that reflect consumer adaptations to COVID. These varied adaptations reveal consumer types based on their reactions. Are these reactions reflecting certain personality traces? With this premise, the paper investigates the potential influence of Big-Five Personalities on the consumer types that emerged out of COVID-19-induced circumstances across the globe. The sample size for this study was one hundred and thirty-four respondents. These personality traits and consumer types were identified and then compared. The consumer preferences and their purchase baskets were then analysed. This study identified four clusters of consumers referred to as consumer types, which were: worrier, individualist, activist and rationalist. No strong associations emerged between personality traits and consumer types, except agreeableness emerging as a common trait across consumer types. There was a similarity in consumer preferences and purchase baskets for all clusters and personality types. The behavioural homogeneity of consumers during this phase of COVID across the globe reassures earlier studies that certain environmental and social situations may moderate or reduce individual differences. [ABSTRACT FROM AUTHOR]

The projection pursuit model is an important tool for processing high-dimensional nonnormal and nonlinear data, and it has a wide range of applications. In this paper, clustering analysis and information entropy theory are simultaneously introduced into the projection pursuit model. The K-means clustering method is used to cluster high-dimensional data, and information entropy is used to measure the overall dispersion and local aggregation of projection data. A projection pursuit model based on clustering and information entropy is proposed. The new model has both the classification advantages of clustering analysis and the evaluation advantages of the projection pursuit model. In the case test, different cases are tested by using the Shapiro-Wilk test, Kolmogorov-Smirnov test, Epps-Pulley test, etc. The results show that in most cases, the new model is better than the original model, and the advantage is clearer when the data dimension is higher. Finally, the new model is applied to the agricultural drought risk assessment of Qiqihar, a major grain-producing area in China that is prone to drought. The regional agricultural drought risk shows a downward trend over time. Spatially, the risk of the central and southern regions is low, while the risk of the northern and western regions is high. Regional ability to resist disasters is the main reason for the spatial and temporal differences. This paper extends projection pursuit model theory, analyzes the characteristics of the spatiotemporal variation in regional agricultural drought risk, and provides a reference for identifying the hidden dangers of drought disasters and disaster reduction. [ABSTRACT FROM AUTHOR]

Cement kiln co-processing (CKC) is a promising technology for solid waste treatment, which can achieve both goals of energy saving for cement production and environment protection for waste reduction, and further development of this technology is desired, so it is necessary to highlight its research gaps between science and technology. In this study, a framework for identifying technological opportunities of cement kiln co-processing was proposed, where (1) scientific and technological keywords from papers and patents were extracted through the RAKE algorithm, and these keywords were divided into four clusters: effect, technology process, object of co-processing and pollutant; (2) the semantic similarities between the cluster science and technology fields were analyzed by calculating their cosine similarity of vectors to determine the technological opportunities, which combined with the occurrence time and frequency of the selected keywords. The results show that the cluster "effect" technologies have been almost developed. In the other three clusters, several technologies are more likely to become technological opportunities for CKC: power generation from recovered heat, co-processing of sludge and raw meal, heavy metal, and NOx emission control. This work provides guidance information for policy research and future studies about CKC. [ABSTRACT FROM AUTHOR]

Deviation of the car from the lane is very dangerous and it leads to crashes. Hence, lane detection is one of the most important features that helps in maintaining the car stay in the respective lane. The main goal of this paper is to detect the lanes with road lane markings with higher accuracy under sharp curve scenarios. In this paper, lane points are extracted using image processing techniques, lane detection of solid lines and dashed lines using a combination of Augmented sliding windows and Clustering technique is discussed. Lanes are simulated in a laboratory for an input data set of 1388 images. The input data set consists of curved lines of dashed and solid lines which split and merge. This combined technique (Augmented sliding windows + clustering) is mainly used to detect the split lanes scenario. Further partially obscured lanes are also tested with the algorithm. The center of the lanes throughout the lane is calculated. Missing lane markings of obscured lanes are also estimated using the relative position of adjacent lanes based on lane width. The detection accuracy of Clustering and Augmented sliding window for the considered input dataset of 1388 frames is 97.70%. The algorithm is showing reliable accuracy in detecting the sharp curves. [ABSTRACT FROM AUTHOR]

Nowadays, the rapid growth of operational costs and human resource shortages increase the importance of the leakage reduction, and the pressure management. This paper presents an optimisation procedure placing pressure-reducing valves (PRVs) to minimise the water losses of real-life water distribution networks (WDNs). The number and placement of the valves are optimised from clustering perspective with the Leiden algorithm. The optimal settings of the valves are determined with a differential evolution algorithm, where a compound fitness function is applied. On the one hand, it minimises the leakages, on the other hand, evades the pressure-dependent consumer outages, i.e. it keeps the pressure above the desired. Seven real-life water distribution networks were analysed from the region of Western Hungary using the techniques. Moreover, three of them are analysed in details. The robustness of the achieved leakage reduction with the optimised PRV placement is analysed to the uncertainty of the leakage exponent. [ABSTRACT FROM AUTHOR]

The directional wave spectrum (DWS) describes the energy of sea waves as a function of frequency and direction. It provides useful information for marine studies and guides the design of maritime structures. One of the challenges in the statistical estimation of DWS is to account for the circular nature of direction. To address this issue, this paper considers the 1-dimensional case of the direction-only DWS (DWSd) and applies the circular regression to smooth the DWSd observations. This paper then improves an existing clustering algorithm by incorporating circular smoothing in the clustering algorithm, automating the determination of the optimal number of clusters, and designing a more appropriate smoothing parameter selection procedure for data with correlated errors. Our simulation studies reveal an improvement in the performance of estimating the underlying DWSd using the circular smoother. Finally, the linear and circular smoothers are compared by clustering two real datasets, one from the Sofar Ocean network and the second from a buoy located at the Red Sea. For the Sofar Ocean data, clustering with the two smoothers results in different number of clusters. For the Red Sea data, a cluster with a peak at the boundary is only identified when the circular smoother is used. Supplementary materials accompanying this paper appear online. [ABSTRACT FROM AUTHOR]

Parkinson disease (PD) is a neurodegenerative disease that occurs due to insufficient level of dopamine in the human brain. This disease may occur predominantly in elder people. There exists no definite procedure to diagnose PD. It is diagnosed based on the symptoms, clinical trials, and number of laboratory tests. In this research paper, machine learning techniques are used to predict PD and help the medical practitioner to recommend personalized drugs for the patients. In this research paper, appropriate features are selected through rough set theory, and principal component analysis is used for dimensionality reduction. The performance is experimented using deep neural network, random forest, and SVM classifiers. The efficiency of the proposed approach is measured through confusion matrix, accuracy, precision, and recall. [ABSTRACT FROM AUTHOR]

Clustered wireless sensor networks (WSNs) are becoming a rapidly emerging technological field. Clustering helps create scalable and robust network infrastructure. It enhances the network lifetime while reducing power consumption. However, clustered WSNs suffer from imbalanced load distribution, inadequate cluster head (CH) allocation and uneven energy consumption issues. In case of improper clustering, nodes get prematurely dead due to excessive energy consumption. This adversely influences the network duration and worsen its performance. To overcome these concerns, this paper attempts to improve the clustering quality through optimal cluster head selection and an advanced cluster formation technique. As this paper tackles the clustering issue in WSNs, it introduces a clustering based routing protocol, named ROA-MOCT. This protocol merges the Raccoon optimization algorithm (ROA) along with a new multi-objective optimization technique to enhance the clustering quality. The ROA algorithm performs an accurate search process to select the optimal CH, based on a fitness function that improves clusters segregation. In addition to CHs selection, ROA-MOCT alters the classical cluster formation process and entrusts it to the CHs. Each CH selects its own cluster members based on a multi-objective optimization technique that reduces the intra-cluster distance. The proposed protocol ROA-MOCT is tested against the conventional LEACH protocol and the recently proposed YSGAP protocol. ROA-MOCT outperforms both LEACH and YSGAP in terms of network lifetime and throughput. [ABSTRACT FROM AUTHOR]

Network densification is a key means which is able to solve the rapid growth of mobile communication traffic. Using millimeter wave with rich spectrum resource to accomplish dense coverage of network has the characteristics of small co-frequency interference and large network capacity. However, it also increases the complexity of resource allocation algorithm and system maintenance becomes inconvenient. In view of this, we propose a clustering algorithm for dense millimeter wave heterogeneous cellular network, which takes the signal-to-interference-plus-noise ratio (SINR) in number form as optimization goal, and completes network clustering according to reference base stations and the maximum number of users within cluster. For the integer programming problem which is difficult to solve after modeling, this paper adopts the extended penalty factor and difference of two convex functions programming to obtain solution. The simulation results show that compared with the clustering method which takes SINR in decibel form as optimization goal, the proposed algorithm has higher stability and practicability because it avoids nonlinear operation. And the algorithm has better network performance than K-means clustering method. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Central South University is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)