Search

At present, the DBSCAN clustering algorithm has been commonly used principally due to its ability in discovering clusters with arbitrary shapes. When the cluster number K is predefined, though the partitional clustering methods can perform efficiently, they cannot process the non-convex clustering and easily fall into local optimum. Thereby the concept of K-DBSCAN clustering is proposed in this paper. But the basic DBSCAN has a crucial defect, that is, difficult to predict the suitable clustering parameters. Here, the well-known harmony search (HS) optimization algorithm is considered to deal with this problem. By modifying the original HS, the novel harmony search (novel-HS) is put forward, which can improve the accuracy of results as well as enhance the robustness of optimization. In K-DBSCAN, the novel-HS is used to optimize the clustering parameters of DBSCAN to obtain better clustering effect with the number of K classifications. Experimental results show that the designed clustering method has superior performance to others and can be successfully considered as a new clustering scheme for further research.

Supervised active contour models can use information extracted from supervised samples to guide contour evolution. However, their applicability is limited by the accuracy of the probabilistic models they use, especially when processing images with intensity inhomogeneity. In this paper, an unsupervised activity contour model with Self Organizing Maps (SOM) is proposed. The proposed model employs the self-organizing neural network to perform clustering calculation and the clustering center is called local self-organizing clustering center. An adaptive sign function is used to control the direction of curve evolution. To improve the stability of curve evolution, an improved double-well potential function is proposed. The experiment results show that our model can effectively segment images with intensity inhomogeneity.

Many machine learning algorithms and models have been proposed in the literature for predicting the remaining useful life (RUL) of systems and components that are subject to condition monitoring (CM). However, in cases where data is ubiquitous, identifying the most suitable equipment for life-extension based on CM data and RUL predictions is a rather challenging task. This paper proposes a technique for determining and prioritizing high-value assets for life-extension treatments when they reach the end of their useful life. The technique exploits the use of key concepts in machine learning (such as data mining and k-means clustering) in combination with an important tool from reliability-centered maintenance (RCM) called the potential-failure (P-F) curve. The RCM process identifies essential equipment within a plant which are worth monitoring, and then derives the P-F curves for equipment using CM and operational data. Afterwards, a new index called the potential failure interval factor (PFIF) is calculated for each equipment or unit, serving as a health indicator. Subsequently, the units are grouped in two ways: (i) a regression model in combination with suitably defined PFIF window boundaries, (ii) a k-means clustering algorithm based on equipment with similar data features. The most suitable equipment for life-extension are identified in groups in order to aid in planning, decision-making and deployment of maintenance resources. Finally, the technique is empirically tested on NASA’s Commercial Modular Aero-Propulsion System Simulation datasets and the results are discussed in detail.

In the past few years, Safe Semi-Supervised Learning (S3L) has become an emerging research topic. A few studies have been investigated in the S3L field and obtained desired performance. However, these studies mainly focus on classification problems which cause less attention on clustering. Meanwhile, there is no study takes both risky labeled and unlabeled samples into consideration(e.g., mislabeled samples and outliers). Therefore, we propose a novel Safe Semi-Supervised clustering method to safely explore the labeled an unlabeled samples. Firstly, we apply an effective approach to compute Safe Degree (SD) by estimating local density and minimum distance of each labeled and unlabeled sample. If a sample has large local density and small minimum distance, it can be safe, and correspondingly SD should be high. Otherwise, the sample should be risky and SD is low. Then the SD is introduced into a model-based semi-supervised clustering method to reduce the negative influences of risky labeled and unlabeled samples. Additionally, we construct a graph-based regularization term to limit the outputs of risky labeled samples to be those of nearest unlabeled neighbors. In this case, it is expected to further reduce the harm of risky labeled samples. At the same time, an illustration on an artificial dataset is given to explain the usefulness of the defined SD. Finally, the results which conducted on ten UCI datasets show that our algorithm is effective enough to achieve good clustering performance.

The segmentation of images under biased conditions such as low contrast, high-intensity inhomogeneity, and noise is a challenge for any image segmentation model. The ideal image segmentation model must be capable of segmenting images with maximum accuracy and a minimum false-positive rate under biased conditions. In this paper, we propose a region-based active contour model (ACM), called global signed pressure and K-means clustering based on local correntropy with the exponential family (GSLCE), to address segmentation challenges under biased conditions. An adaptive weighted function is formulated based on the global and local image differences such that a single weighted function can drive both the global and local intensities. Further, the Riemannian steepest descent method is used for convergence of the proposed GSLCE energy function, and a Gaussian kernel is applied for spatial smoothing to obviate the computationally expensive level-set re-initialization. The experimental results show that, compared with state-of-the-art ACMs, the proposed GSLCE model obtained the best visual image segmentation results for synthetic and real images under biased conditions. Further, the qualitative and quantitative experimental results validate that the proposed model outperforms the state-of-the-art ACMs by yielding higher values of performance metrics. Moreover, the proposed GSLCE model requires substantially less processing time compared to the state-of-the-art ACMs.

K-Means algorithm is probably the most famous and popular clustering algorithm in the world. K-Means algorithm has the advantages of simple structure, easy implementation, high efficiency, fast convergence speed, and good results. It has been widely used in many applications, and many extensions of K-Means have been proposed. Basically, most K-Means variants deal with static data. Recently, the dynamic nature of data has received increasing attention from researchers. Therefore, some studies also use K-Means algorithm to deal with clustering problems in evolutionary data. In this article, we aim to improve past variants of K-Means used in evolutionary clustering. There are two ways to improve this problem. First, past research only considered how the previous clustering results affected the current clustering, but we also considered how the future clustering results affect the current clustering. Secondly, past research applied K-Means from one cycle to another in one pass, but we extended it to multiple passes. These two improvements make the proposed algorithm MK-Means provide more consistent, stable and smooth clustering results than previous models.

Multi-view clustering has attracted increasing attention in recent years since many real data sets are usually gathered from different sources or described by different feature types. Amongst various existing multi-view clustering algorithms, those that are based on non-negative matrix factorization (NMF) have exhibited superior performance. However, NMF decomposing original data directly fails to exploit global relationships between data samples and cannot be applied to datasets that are not strictly non-negative. In this paper, a network-based sparse and multi-manifold regularized multiple NMF (NSM_MNMF) for multi-view clustering is proposed, where multi-view data is transformed into multiple networks, and NMF is used to jointly factorize transformed multiple networks for capturing the shared cluster structure embedded in different views. Furthermore, sparse and multi-manifold regularization are incorporated into NMF to keep the intrinsic geometrical information of the multi-view network manifold space. Networks characterize intra-view similarity, and joint factorization reveals inter-view similarity across distinct views, while using NMF to decompose the networks instead of the original data means NSM_MNMF can be applied to datasets that are not strictly non-negative and the clustering results are interpretable. Extensive experiments are conducted on nine real data sets to assess the method proposed, and the results illustrate that NSM_MNMF outperforms other baseline approaches.

The tremendous growth in the technology has led to the accumulation of enormous Big Data. Techniques that efficiently analyse this Big Data are in great demand. Tweets from Social media and Sensor data are some of the most common forms of Big Data. Machine learning algorithms pave way for researchers to analyze Big Data. Most Machine learning algorithms depend on efficient feature extraction and feature selection for its success. Here, we explore feature selection methods like entropy and Rough set on the sensor data. Also a symbolic approach of feature extraction is proposed which represents both sensor and twitter data efficiently for further data analysis. Some popular classifiers like Naive Bayes, K Nearest Neighbour, Support Vector Machine and Decision Tree are used for validating the efficacy of the features selected. An ensemble classifier technique is also proposed which is compared with various state of the art ensemble classifiers. Symbolic features perform better than both entropy and Rough set features for sensor data and improves the clustering efficiency of twitter data. The proposed ensemble weighted average classifier on Symbolic features outperform all the other ensemble classifiers and independent classifiers. The results obtained from these methods have the potential to aid the public health surveillance.

Given a network, local community detection (a.k.a. graph clustering) methods aim at finding communities around the selected initial nodes (also referred to as seeds, starting nodes or core nodes). Methods in this kind successfully address the efficiency problem confronted by global clustering methods. And techniques, such as personalized PageRank and heat kernel diffusion, for ranking the proximity score of vertices nearby with respect to the corresponding starting nodes are developed. However, most of the random-walk based metrics allow a walker to diffuse without any constraint, and the walker can easily run into irrelevant communities. As a result, the corresponding community could include irrelevant high-quality communities (communities with good fitness score) nearby, we refer to the case that a walker goes into irrelevant communities and causes inaccurate expansion of a community as redundant diffusion. In this work, we develop a constrained personalized PageRank method for community expansion to reduce the problem of redundant diffusion. In the mechanism, a walker moves with lower probability to neighbor nodes already in the existing communities, and a walker tends to walk out of the community if the walker walks into an irrelevant community. Extensive experiments on synthetic and large real-world networks demonstrate that the proposed method outperforms approaches in the state of the art by a large margin in accuracy and efficiency.

The key challenges for automatic hate-speech classification in Twitter are the lack of generic architecture, imprecision, threshold settings and fragmentation issues. Most studies used binary classifiers for hate speech classification, but these classifiers cannot really capture other emotions that may overlap between positive or negative class. Hence, a probabilistic clustering model for hate speech classification in twitter was developed to tackle problems with hate speech classification. A metadata extractor was used to collect tweets containing hate speech keywords and a crowd-sourced experts was employed to label the collected hate tweets into two categories: hate speech and non-hate speech. Features representation was done with Term Frequency- Inverse Document Frequency (TF-IDF) model and enhanced with topics inferred by a Bayes classifier. A rule-based clustering method was used to automatically classify real-time tweets into the correct topic clusters. Fuzzy logic was then used for hate speech classification using semantic fuzzy rules and a score computation module. From the evaluation results, it was observed that the developed model performed better in hate speech detection with F1-sore of 0.9256 using a 5-fold cross validation. Similarly, the developed model for hate speech classification performed better with F1-score of 91.5 compared to related models. The developed model also indicates a more perfect test having an AUC of 0.9645, when compared to similar methods. The Paired Sample t-Test validated the efficiency of the developed model for hate speech classification.

Document clustering in text mining is a problem that is heavily researched upon. It is observed that individual approaches based on statistical features and semantic features have been extensively used to solve this problem. However, techniques combining the advantages of both types of features have not been frequently researched upon. Specifically, when the growth in the size of textual data is immense, there is a need for such an approach that combines the advantages of both types of features to give more accurate results within an acceptable range of time. In this paper, a document clustering technique is proposed that combines the effectiveness of the statistical features (using TF-IDF) and semantic features (using lexical chains). It is designed to use a fewer number of features while maintaining a comparable and even better accuracy for the task of document clustering.

Internet of Medical Things (IoMT) is a recently introduced paradigm which has gained relevance as an emerging technology for widely connected and heterogeneous networks. In the medical context, these networks involve many different processes run by different types of devices called objects that interact and collaborate to achieve a common goal (e.g, diagnosis, treatment, monitoring, or rehabilitation of a patient). An IoMT framework in a smart healthcare system dynamically monitors patients to respond to their assistance demands so that vital signs of critical or unusual cases can be uncovered based on the collected data. To this end, an effective technique called SIoMT (Swarm Intelligence optimization technique for the IoMT) is proposed in this paper for periodically discovering, clustering, analyzing, and managing useful data about potential patients. Notably, the SIoMT technique is widely used with distributed nodes for analyzing and managing data groups. Different from the existing clustering algorithms, SIoMT performs clustering based on the characteristics and distance between objects or swarms. More specifically, these data are collected and grouped, in early stage, using a clustering approach inspired by the Bee Colony Optimization algorithm (BCO), adopting some standard quality measures which helped minimizing the latency and required computational cost. To test the performance of the proposed SIoMT, one public dataset (Ward2ICU) was considered from the online source. Various experiments were done to analyze the effects of different parameters on the proposed SIoMT’s performance, and the results from the final variant of the proposed algorithm were compared against different variants of the same algorithm with different clustering algorithms and different optimization algorithms. Subsequently, after analyzing different components by solving various IoMT datasets, the capability and the superiority of the proposed SIoMT approach is well-established among its competitive counterparts.

Clustering is a powerful technique for providing class labels of data objects for learning guidance. Because traditional clustering methods rarely consider the component correlations of a multivariate time series(MTS), an MTS clustering method based on a component relationship network (CRN) is proposed in the present study. An MTS dataset is mapped to a multi-relationship network (MRN), which consists of a set of CRNs. Every CRN reflects the relationship of the MTS data under each component. The proposed method applies two steps. First, the distance function and K-nearest neighbors are combined to transform an MTS dataset into an MRN. Second, a non-negative matrix factorization is designed to identify time-series clusters. Because asynchronous time-series data exist in the form of an approximate case, we recommend an improved penalty-coefficient based dynamic time-warping algorithm to measure the similarity between two-time sequences. The similarity can reflect the correlation between the asynchronous MTS data. Through an experiment, we compared the proposed method to other clustering methods and discussed of the effects of the parameters in detail. The numerical results of the experiment demonstrate that the proposed method can improve the accuracy and quality of MTS data clustering.

This study introduces a pre-processing step for spectral clustering.A parameter-free similarity graph, Density Adaptive Neighborhood (DAN), is proposed.DAN works on the data sets with arbitrary shaped clusters and density variations.DAN is robust to the number of attributes, geometric distortion and decimation.DAN facilitates the use of spectral clustering algorithms in various domains. Spectral clustering is a popular clustering method due to its simplicity and superior performance in the data sets with non-convex clusters. The method is based on the spectral analysis of a similarity graph. Previous studies show that clustering results are sensitive to the selection of the similarity graph and its parameter(s). In particular, when there are data sets with arbitrary shaped clusters and varying density, it is difficult to determine the proper similarity graph and its parameters without a priori information. To address this issue, we propose a parameter-free similarity graph, namely Density Adaptive Neighborhood (DAN). DAN combines distance, density and connectivity information, and it reflects the local characteristics. We test the performance of DAN with a comprehensive experimental study. We compare k-nearest neighbor (KNN), mutual KNN, e-neighborhood, fully connected graph, minimum spanning tree, Gabriel graph, and DAN in terms of clustering accuracy. We also examine the robustness of DAN to the number of attributes and the transformations such as decimation and distortion. Our experimental study with various artificial and real data sets shows that DAN improves the spectral clustering results, and it is superior to the competing approaches. Moreover, it facilitates the application of spectral clustering to various domains without a priori information.

With the development of location-aware technology, a large amount of location data of users is collected by the trajectory database. If these trajectory data are directly used for data mining without being processed, it will pose a threat to the user's personal privacy. At the moment, differential privacy is favored by experts and scholars because of its strict mathematical rigor, but how to apply differential privacy technology to trajectory clustering analysis is a difficult problem. To solve the problems in which existing trajectory privacy-preserving models have poor data availability or difficulty to resist complex privacy attacks, we devise novel trajectory privacy-preserving method based on clustering using differential privacy. More specifically, Laplacian noise is added to the count of trajectory location in the cluster to resist the continuous query attack. Then, radius-constrained Laplacian noise is added to the trajectory location data in the cluster to avoid too much noise affecting the clustering effect. According to the noise location data and the count of noise location, the noise clustering center in the cluster is obtained. Finally, it is considered that the attacker can associate the user trajectory with other information to form secret reasoning attack, and secret reasoning attack model is proposed. And we use the differential privacy technology to give corresponding resistance. Experimental results using the open data show that the proposed algorithm can not only effectively protect the private information of the trajectory data, but also ensure the data availability in cluster analysis. And compared with other algorithms, our algorithm has good effect on some evaluation indicators.

Unsupervised feature selection (UFS) is utilized in various application domains, such as data mining, pattern recognition, machine learning, etc. UFS follows three basic approaches, namely filter, wrapper, and hybrid (that is, a combination of both filter and wrapper) to select the relevant and non-redundant features. It has been observed that a filter method does not guarantee an optimal solution. However, a wrapper approach is computationally expensive. The hybrid method are known to give a better trade-off between filter and wrapper strategies. But, the practical applicability of schemes mentioned above are preferably restricted only to a numerical dataset and are not so suitable for a mixed dataset. Therefore, there is a need for a UFS scheme which can handle both the numerical and non-numerical features directly. In this paper, a robust and efficient two-phase (i.e., feature ranking (FR) and feature selection (FS)) UFS method is proposed. The proposed FR utilizes entropy and mutual information to produce maximum informative and non-redundant ranked features from a high-dimensional mixed dataset. Further, the proposed FS follows k-prototype clustering algorithm with improved Callinski-Harasbaz criteria-based selection methodology to choose optimal features. Experiments on real-life dataset substantiate that the proposed approach provides a better subset of features compared to the existing state of the art approaches.

Large-scale group decision-making (LSGDM) problems generally involve a large number of decision-makers (DMs). In many situations, the number of DMs and alternatives simultaneously make traditional techniques inoperable. This paper proposes a two-stage subgroup decision-making (T-SSGDM) method. In contrast to the traditional LSGDM approaches, this framework does not need a clustering process to reduce the size of DMs/alternatives to a manageable level. Instead, we propose a T-SSGDM process to address inter-group heterogeneity. First, DMs and alternatives are randomly grouped so that the number of alternatives to be assessed by each DM is substantially reduced. Second, as the ratings obtained in the first stage of the decision-making process are incomparable, partial samples are selected for the second stage. The relationships among the ratings of different subgroups are then determined by applying the equivalence test. Additionally, to ensure the robustness of the results, three sampling methods and three kinds of functions for the equivalence test are implemented. Finally, an empirical application is used to verify the effectiveness of the proposed method.

Multidimensional data sets are becoming more frequent in practically all research fields, and require complex data analysis techniques in order to extract knowledge from them. In this paper, we propose an interactive visualization tool for performing exploratory data analysis. The tool combines unsupervised and supervised dimensionality reduction methods, such as linear discriminant analysis, or t-SNE, with clustering and classification techniques. Analysts can use several machine learning methods for extracting data structure, and can group data into clusters interactively or through clustering algorithms. In addition they can visualize projections of the data to evaluate the quality of obtained clusters, and to analyze the performance of classification methods. We have applied this tool to analyze a clinical data set related to patients with dermatologic conditions that are under photodynamic therapy. The analysis allowed medical doctors to identify several clinically interesting patient groups. In addition, clinicians discovered a greater efficacy in the treatment of patients with the photosensitizer 5-aminolaevulinic acid nanoemulsion gel compared to those treated with methyl-5-aminolaevulinate cream.

Humans must easily handle the vast amounts of data being generated by the revolution of information technology. Thus, Automatic Text summarization has been applied to various domains in order to find the most relevant information and make critical decisions quickly. In the context of Arabic, text summarization techniques suffer from several problems. First, most existing methods do not consider the context or domain to which the document belongs. Second, the majority of the existing approaches are based on the traditional bag-of-words representation, which involves high dimensional and sparse data, and makes it difficult to capture relevant information. Third, research in Arabic Text summarization is fairly small and only recently compared to that on Anglo-Saxon and other languages due to the shortage of Arabic corpora, resources, and automatic processing tools. In this paper, we try to overcome these limitations by proposing a new approach using documents clustering, topic modeling, and unsupervised neural networks in order to build an efficient document representation model. First, a new document clustering technique using Extreme learning machine is performed on large text collection. Second, topic modeling is applied to documents collection in order to identify topics present in each cluster. Third, each document is represented in a topic space by a matrix where rows represent the document sentences and columns represent the cluster topics. The generated matrix is then trained using several unsupervised neural networks and ensemble learning algorithms in order to build an abstract representation of the document in the concept space. Important sentences are ranked and extracted according to a graph model with a redundancy elimination component. The proposed approach is evaluated on Essex Arabic Summaries Corpus and compared against other Arabic text summarization approaches using ROUGE measure. Experimental results showed that the models trained on topic representation learn better representations and improve significantly the summarization performance. In particular, ensemble learning models demonstrated an important improvement on Rouge recall and promising results on F-measure.

The chapter introduces a consensus model for group decision making with hesitant fuzzy linguistic preference relations (HFLPRs) studied in [7], which involves clustering HFLPRs based on a kernel function, and establishing a consensus model based on measuring the modified extents of decision makers’ HFLPRs. Experiments are furtherly shown to discuss the convergence of the proposed consensus model. Finally, the chapter provides the illustration of the model through a case study in choosing a flood discharge technique for a hydropower station.

Medical diagnosis is a challenging procedure that involves issues such as data imbalance, insufficient labels, obscure images, redundancy and lack of effective model training directions to shrink the semantic gap between human knowledge and computer algorithms. Due to privacy norms, sometimes medical images are difficult to access and therefore, retrieval of identical existing images from an existing repository is quite useful. This paper proposes a search space to narrow down the identical images in an archive by using (i) Capsule Networks, followed by a (ii) decision fusion with Wavelet-Discrete Cosine Transform (W-DCT) and Radon Barcodes (RBC). Empirical case study has been applied on IRMA (Image Retrieval in Medical Applications) dataset, ImageCLEFMed-2009, containing 14,410 X-ray images, but the proposed method is generic, reproducible and scalable. Subjective and quantitative performance has been compared with the state-of-art and it has been found superior to yield accuracy of 92.83% and IRMA error of 124.25 for 193 class-code category. Thus, the proof-of-concept helps to improves diagnosis efficiency for automatic image retrieval and annotation by clustering similar images from the underlying repository.

The spectral clustering algorithm is extensively employed in different aspects, especially in the field of pattern recognition. However, the efficient construction of the neighborhood graph is the main reason for its promising results. Generally, the similarity matrix relies on the applied similarity measure between two data points, selection of k − nearest neighbors (KNN), and approach for the construction of a neighborhood graph. In this study, we integrate S-distance to spectral clustering, which is capable to find out the complex and non-linear cluster structures. Moreover, generalized mean distance-based KNN is proposed to decrease the sensitiveness towards the value of the k . Also, a symmetry-favored KNN method is applied to construct the neighborhood graph, which reduces the impact of outliers and noisy data points. However, spectral clustering faces scalability and speedup issues in the case of large size datasets. Thus, the proposed spectral clustering algorithm is also executed in distributed environments. Several experiments are performed to validate the proposed clustering algorithm on 20 real-world datasets and 3 large size datasets. Experimental results demonstrate that the proposed clustering algorithm outperforms some of the baseline methods in terms of accuracy and clustering error rates. Finally, we conduct Wilcoxon’s Rank-Sum test and illustrate that the proposed spectral clustering algorithm is statistically significant.

Sentiment Analysis is a process that aids in assessing the performance of products or services from user generated online posts. In present time, there are various websites that allow customers to post reviews about movies, products, events or services, etc. This has led to cumulative aggregation of a lot of reviews written in natural language. Prevailing factors such as availability of online reviews and raised end-user expectations have motivated the evolution of opinion mining systems that can automatically classify customers' reviews. It is observed that in Sentiment Analysis (SA), to highlight the significant keyphrases which contribute towards correct sentiment cognition is a tedious task. In this paper, we have proposed an unsupervised sentiment classification system that comprehensively formulates phrases, computes their senti-scores (sentiment scores) and polarity using the SentiWordNet lexicon and fuzzy linguistic hedges. Further it extracts the keyphrases significant for SA using fuzzy entropy filter and k-means clustering. We have deployed document level SA on online reviews using n-gram techniques, specifically combination of unigram, bigram and trigram. Experiments on two benchmark movie review datasets- polarity dataset by Pang and Lee and IMDB dataset, achieve high accuracy for our approach as compared to the other state-of-the-art-methods for phrase-level SA.

In this paper, we address an online dimensionality reduction approach to deal with a many-objective formulation of a Vehicle Routing Problem with a Demand Responsive Transport (VRPDRT). The problem relates to a mode of transport similar to available carpooling services in which passengers are transported from their origin to their destination sharing the same vehicle. The goal is to reduce the operating/riding costs while meeting passenger needs and providing high-quality service. Due to the complexity and conflicting characteristics of the problem, an evolutionary approach based on a dimensionality reduction technique is applied to solve the VRPDRT in which eight different objective functions are used. The performance of the proposed approaches – OnCLτ-MOEA/D and OnCLρ-MOEA/D – are compared to an a priori cluster dimensionality reduction with Pearson’s and τ of Kendall correlation coefficients using a realistic data set containing distances and travel time for Belo Horizonte, Brazil. The online and offline versions of the algorithms are also compared with a baseline approach, a classic version of MOEA/D. Results show that the online cluster-based approach achieves a better spread of solutions, when compared to its a priori versions. Moreover, there is no difference between the results obtained from the online Cluster-based approach and the original MOEA/D. It shows that the proposed dimensionality reduction is an effective technique presenting a positive effect on the search efficiency, computational cost, and in the application of usual visualization techniques.

Collaboration among logistics companies offers a simple and effective way of increasing logistics operation efficiency. This study designs an optimal collaboration strategy by solving the collaborative logistics pickup and delivery problem with eco-packages (CLPDPE). This problem seeks to minimize the total operational costs by forming collaborative alliances and allocating trucking resources based on time–space (TS) network properties. The synchronization of two-echelon logistics networks is improved by solving this problem. Moreover, this study considers the stability of collaboration (i.e., the willingness of logistics companies to join and remain in collaborative alliances) by comparing different profit allocation strategies in the CLPDPE solving process. A novel methodology that combines multi-objective mixed integer programming, multidimensional K-means clustering, reference point based non-dominated sorting genetic algorithm-II (RP-NSGA-II), forward dynamic programming and improved Shapley value method is developed to formulate and solve CLPDPE. Our results show that the proposed algorithm outperforms most other algorithms in minimizing the total cost, waiting time and number of vehicles. An empirical case study in Chongqing city, China suggests that the proposed collaborative mechanism and transportation resource sharing strategy based on TS network can reduce cost, improve distribution efficiency, and contribute to efficient, smart, intelligent and sustainable urban logistics and transportation systems.

Gaussian kernel is defined by Euclidean distance and has been widely used in many fields. In the view of Euclidean distance is sensitive to outliers or noise and it is difficult to obtain satisfactory results for complex non-convex data. Entropy-like divergence is firstly induced by combining Jenson-Shannon/Bregman divergence with convex function, and its mercer kernel function called entropy-like divergence-based kernel is also constructed in this paper. Secondly, a new fuzzy weighted factor based on entropy-like divergence-based kernel is proposed by improving the existing trade-off weighting factor of kernel-based fuzzy local information C-means clustering (KWFLICM). In the end, a weighted fuzzy local information clustering based on entropy-like divergence-based kernel (EKWFLICM) is presented, which combines a new weighted fuzzy factor and entropy-like divergence-based kernel. Experiment results show that the proposed algorithm outperforms the segmentation performance of existing state-of-the-art fuzzy clustering-related algorithms for the image in presence of high noise.

Automated mammogram image segmentation is one of the most important methods in the domain of medical diagnosis and decision systems. Accurate segmentation of mammogram plays a key role for the detection of any kind of abnormality like lesion tissues, cyst in mammogram images for medical diagnosis. In this study, a novel hybrid soft computing entrenched segmentation method for mammograms is introduced for the detection of breast cancer in early stages. Here, we have designed a novel automatic mammogram segmentation method using intuitionistic fuzzy soft sets (IFSS) and Multigranulation rough set. First, the proposed clustering algorithm accomplishes a soft information structure from the source image using IFSSs via multiple fuzzy membership functions with Yager generating function. The IFSS handles the ambiguity among lesion and non-lesion pixels through the hesitant degree while shaping the membership function. To reduce distant pixels which do not belong to the region of interest (ROI), the lesion tissues in mammogram image is segregated by decision making scheme via a rough approximation of a fuzzy concept under the field of multigranulation space. Later, the proposed scheme utilizes soft-information builder with accuracy and roughness scores on multigranulation approximation space for segregation of normal and abnormal (lesion tissues) pixel from mammograms. The proposed model has generated a threshold image from accuracy and roughness scores via defuzzification process. The proposed segmentation model performs better than the existing methods using evaluation metrics like segmentation accuracy, Jeccards similarity coefficient.

In recent years, the rapid development of mobile network and wireless sensor technology has brought opportunities to change the way of the existing taxi business operation. How to improve the operation revenues of taxi drivers has become a topic worthy of research. This paper analyzes and mines taxi operation data to provide taxi drivers with personalized sequence recommendation services, thereby increasing their expected revenues. Different from previous works, the proposed method in this paper recommends a series of future operation orders for taxi drivers, instead of recommending several discrete locations for the current order. In this paper, firstly, by performing spatial-temporal clustering on the origins and destinations of passengers, the spatial and temporal distribution characteristics of passengers in the city are identified. Secondly, the origin of the current passenger is used as the root node to construct a top-N high-utility sequence tree, and this process can be divided into two processes: top-down building tree and bottom-up sorting path utility. The two pruning strategies of node utility and path utility are used to reduce the generation of candidate sets. Finally, a series of potential orders based on dynamic context are recommended to taxi drivers, so as to maximize the expected revenues of taxi drivers. The experimental results demonstrate that there is a close relationship between taxi drivers’ operation behavior patterns and their revenues. The proposed system framework and algorithm in this paper can effectively mine global and long-term top-N high-utility operation patterns.

Mixed datasets consist of both numeric and categorical attributes. Various k-means-based clustering algorithms have been developed for these datasets. Generally, these algorithms use random partition as a starting point, which tends to produce different clustering results for different runs. In this paper, we propose, initKmix, a novel algorithm for finding an initial partition for k-means-based clustering algorithms for mixed datasets. In the initKmix algorithm, a k-means-based clustering algorithm is run many times, and in each run, one of the attributes is used to create initial clusters for that run. The clustering results of various runs are combined to produce the initial partition. This initial partition is then used as a seed to a k-means-based clustering algorithm to cluster mixed data. Experiments with various categorical and mixed datasets showed that initKmix produced accurate and consistent results, and outperformed the random initial partition method and other state-of-the-art initialization methods. Experiments also showed that k-means-based clustering for mixed datasets with initKmix performed similar to or better than many state-of-the-art clustering algorithms for categorical and mixed datasets., Comment: 41 page, 3 Figures, 19 Tables, 5 Algorithms

In recent years, the size of graph data has increased significantly, but most existing graph clustering algorithms do not consider the case where the size of main memory is not sufficient to handle large amount of graph data. Exploring entire region of graph for clustering causes too many random disk accesses to use data that are not loaded into memory, resulting in excessive disk I/O and thrashing. To address this problem, we propose an I/O-efficient algorithm for structural clustering of a graph, called pm-SCAN. In the proposed method, if memory is insufficient, an input graph is partitioned into several subgraphs smaller than memory, and clustering is first performed for each subgraph. And then clusters from the subgraphs are merged based on connectivity between clusters so that global results can be obtained in the point of view of an original input graph. Not only does pm-SCAN produce scalable performance even for very large graphs, i.e., significant shortage of available memory, but also the result of pm-SCAN is the same as that of the original structural clustering algorithm SCAN. We also propose a cluster maintenance method for large-scale dynamic graphs that change over time. Instead of reclustering with a whole graph, only a small set of nodes whose structural connectivities are subject to change by a given update operation is first identified, and we access only those nodes in disk and update their clusters to reduce maintenance costs. This dynamic graph handling mechanism shows significant performance improvement compared to the existing method and the baseline that performs clustering from scratch.

K-means is a well-known unsupervised-learning algorithm. It assigns data points to k clusters, the centers of which are termed centroids. However, these centroids have a structure usually represented by a list of quantized vectors, so that kernel density estimation models can better represent complex data distributions. This paper proposes a k-means-based supervised-learning clustering method termed supervised kernel-density-estimation k-means. The proposed approach uses kernel density estimation for class examples inside each cluster to obtain a better representation of the data distribution. The algorithm constructs an initial model using supervised k-means with an equal seed distribution among the classes so that a balance between majority and minority classes is achieved. We also incorporate incremental semi-supervised learning into the proposed method. Experiments were conducted using publicly available benchmark datasets. The results demonstrated that, compared with state-of-the-art supervised methods, the proposed algorithm, which can also perform incremental semi-supervised learning, achieved highly satisfactory performance.

Collaboration among logistics operators offers an effective way to improve customer service and freight transportation efficiency. One form of collaboration is the sharing of logistics resources (e.g., delivery vehicles). Existing studies on collaboration and resource sharing have not sufficiently accounted for the time frame within which collaboration happens, and they mostly assume that collaboration among logistics operators occurs in a single time period. This study addresses the issue of collaboration across multiple time periods, in which logistics resources can be shared between different service time periods, by formulating and solving a two-echelon collaborative multi-depot multi-period vehicle routing problem (2E-CMDPVRP). The 2E-CMDPVRP is formulated as a multi-objective integer programming model that minimizes logistics operational costs, service waiting times, and number of vehicles in multiple service periods. A hybrid heuristic algorithm with three-dimensional k-means clustering and improved reference point-based non-dominated sorting genetic algorithm-III (IR-NSGA-III) is proposed to solve the multi-objective optimization model. Comparative analysis results show that the proposed IR-NSGA-III outperforms existing algorithms in terms of the minimization of logistics operational costs, service waiting times, and number of vehicles. The minimum costs remaining saving method and strictly monotonic path selection principle are combined to determine the best profit allocation schemes and the optimal coalition sequences. An empirical case study of a multi-depot multi-period logistics network in Chongqing, China, is used to validate the proposed model and solution algorithm. Results suggest that the proposed collaborative mechanism with multi-depot and multi-period resource sharing can improve the degree of synchronization within a collaborative logistics network, and thus contribute to sustainable development of urban logistics distribution networks.

Search Results Clustering (SRC) is a well-known problem in the field of information retrieval and refers to the clustering of web-snippets for a given query based on some similarity/dissimilarity measure. In this current study, we have posed Search Results Clustering problem as a multi-view clustering problem and solved it from an optimization point of view. Various views based on syntactic and semantic similarity measures were considered while performing the clustering. In contrast to existing algorithms, three new views based on word mover distance, textual-entailment, and universal sentence encoder, measuring semantics while performing clustering, are incorporated in our framework. Different quality measures computed on clusters generated by different views are optimized simultaneously using multi-objective binary differential evolution (MBDE) framework. MBDE comprises a set of solutions and each solution is composed of two parts corresponding to different views. An agreement index checking the accordance between partitionings of different views is also optimized to obtain a consensus partitioning. The proposed approach is automatic in nature as it is capable of detecting the number of clusters for any query in an automatic way. Experiments are performed on three benchmark multi-view datasets corresponding to web search results and evaluated using well-known F-measure metric. Results obtained illustrate that our approach outperforms state-of-the-art techniques.

Clustering is the process of identifying objects with similar intrinsic properties and grouping them together into separate clusters. It also serves as a preliminary step in data classification by separating heterogeneous data into reasonably homogeneous groups, which can be further processed. Existing literature discusses a host of different clustering algorithms and their applications, albeit no single approach for all applications has yet emerged. In this paper, we introduce a novel clustering algorithm, YAC2, based on data binning and α-proximity/neighborhood. The binning process is a data transformation step that converts cardinal values of the attributes into their ordinal equivalence. The algorithm introduces a specialized centroid that is used with the α-proximity and a matching algorithm to partition data set into a sequentially generated clusters. We present the results of applying YAC2 to a set of established benchmark data sets, using a host of evaluation metrics. Our results show YAC2 to perform well besting a number of well-established algorithms. For several metrics, YAC2 has provided improvements averagely in the range of 6% (in the Iris data set)-180% (in the Wholesale Customer data set) over other algorithms.

In digital forensics, user profiling aims to predict characteristics of the user from digital evidence extracted from digital devices (e.g. smartphone, laptop, tablet). Previous researches showed promising results, but there are limitations to apply practical investigations. The researches so far have focused only on specific applications, devices, or operating systems by analyzing the order of execution or volatile data such as network traffic and online content. This paper introduces a user profiling method, named Entity Profiling with Binary Predicates (EPBP) model, which analyzes non-volatile data remained on digital devices. The proposed model defines that a user has two properties: tendency and impact, which indicate patterns of application usage. Based on the attributes, the EPBP model generates users’ profiles and performs similarity analysis to differentiate between the users. We also present methods for clustering and anomaly detection through real case studies.

Collaborative filtering (CF) recommendation is well-known for its outstanding recommendation performance, but previous researches showed that it could cause privacy leakage for users due to k -nearest neighboring (KNN) attacks. Recently, the notion of differential privacy (DP) has been applied to privacy preservation in recommendation systems. However, as far as we know, existing differentially private CF recommendation systems degrade the recommendation performance (such as recall and precision) to an unacceptable level. In this paper, to address the performance degradation problem, we propose a differentially private user-based CF recommendation system based on k -means clustering (KDPCF). Specifically, to improve the recommendation performance, KDPCF first clusters the dataset into categories by k -means clustering and appropriately adjusts the size of the target category to which the target user belongs, so that only users in the well-sized target category are used for recommendation. Then, it selects efficiently a set of neighbors from the target category at one time by employing only one instance of exponential mechanism instead of the composition of multiple ones, and then uses a CF algorithm to recommend based on this set of neighbors. We theoretically prove that our system achieves differential privacy. Empirically, we use two public datasets to evaluate our recommendation system. The experimental results demonstrate that our system has a significant performance improvement compared to existing ones.

The class imbalance problem occurs when one class far outnumbers the other classes, causing most traditional classifiers perform poorly on the minority classes. To tackle this problem, a plethora of techniques have been proposed, especially centered around resampling methods. This paper introduces a two-stage method that combines the DBSCAN clustering algorithm to filter noisy majority class instances with a graph-based procedure to overcome the class imbalance. We then experimentally evaluate the behavior of the proposed method on a collection of two-class imbalanced data sets. The experimental results show an improvement in the classification performance measured by the geometric mean of the accuracy on each class and also a higher reduction in the imbalance ratio when compared to several state-of-the-art under-sampling techniques.

A citation network technology named main path analysis is used in this study, which provides a historical perspective of Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS). The citations between related papers are regarded as edges in social network and the corresponding weights are allocated according to their role in knowledge diffusion. Several different main paths are implemented in this work to investigate the knowledge structure of TOPSIS. The Louvain clustering algorithm is conjoined with main path analysis to present the development trend in several scientific communities. The visualization of multiple main paths shows the overall knowledge structure instead of a single development trajectory. This is the first article to use this unique method to process such large-scale data in the TOPSIS domain, which provides an insightful view of TOPSIS publications for future research.

A new methodology to predict one-day-ahead hourly solar global radiation is proposed in this paper. This information is very useful to address many real problems; for instance, energy-market decision making is one of the contexts where that information is essential to ensure the correct integration of grid-connected photovoltaic solar systems. The developed methodology is based on the contribution of different experts to obtain improved data-driven models when included in the data mining process. The modelling phase, when models are induced and new patterns can be identified, is the one that most benefits from that expert knowledge. In this case, it is achieved by combining clustering, regression and classification methods that exploit meteorological data (directly measured or predicted by weather services). The developed models have been embedded in a prediction system that offers reliable forecasts on next-day hourly global solar radiation. As a result of the automatic learning process including the knowledge of different experts, 14 different types of day were identified based on the shape of hourly solar radiation throughout a day. The conventional definitions of types of days, that usually consider 4 options, are updated with this new proposal. The next-day prediction of hourly global radiation is obtained in two phases: in the first one, the next-day type is obtained from among the 14 possible types of day; in the second one, values of hourly global radiation are obtained using the centroid of the predicted type of day and extraterrestrial solar radiation. The relative root mean square error of the prediction model is less than 20%, meaning a significant reduction compared to previous models. Moreover, the proposed models can be recognized in the context of eXplainable Artificial Intelligence.

Clustering algorithms are affected by the initial seeds, therefore any improvement of the initialization process can improve the final clustering results. There exist several initialization algorithms that most of them are focused on using the distance and density based metrics defined in the feature space. However image space has a great potential to be used as the search space for initial seeds. In this research, developing clustering initialization using spatial information (image space) and spectral information (feature space) with the help of particle swarm optimization has been examined. Standard deviation and homogeneity of pixels in the image space in addition to distance and density of points in the feature space have been utilized in the objective function of the particle swarm optimization. Two different search spaces (feature and image spaces) and 26 objective functions have been applied to a simulated image and two real satellite multi-spectral images. Comparing the results of 26 cases with four prevailing initialization methods, demonstrated that searching for initial seeds in the image space using PSO with a full objective function (using four spectral–spatial criteria) can produce better results than the other tested cases. Using this case for k-means clustering initialization, led to about 20% improvement in overall accuracy relative to the clustering results with commonly used initialization algorithms.

A novel and efficient fingerprint indexing based on minutiae triplets, is proposed.The proposed fingerprint features are invariant to rotation and translation.Our fingerprint representation is very robust to distortions and noise.The experimental results demonstrate the validity of the proposed algorithm. Fingerprint indexing plays a key role in the automatic fingerprint identification systems (AFISs) which allows us to speed up the search in large databases without missing accuracy. In this paper, we propose a fingerprint indexing algorithm based on novel features of minutiae triplets to improve the performance of fingerprint indexing. The minutiae triplet based feature vectors, which are generated by ellipse properties and their relation with the triangles formed by the proposed expanded Delaunay triangulation, are used to generate indices and a recovery method based on k-means clustering algorithm is employed for fast and accurate retrieval. The proposed expanded Delaunay triangulation algorithm is based on the quality of fingerprint images and combines two robust Delaunay triangulation algorithms. This paper also employs an improved k-means clustering algorithm which can be applied over large databases, without reducing the accuracy. Finally, a candidate list reduction criteria is employed to reduce the candidate list and to generate the final candidate list for matching stage. Experimental results over some of the fingerprint verification competition (FVC) and national institute of standards and technology (NIST) databases show superiority of the proposed approach in comparison with state-of-the-art indexing algorithms. Our indexing proposal is very promising for the improvement of real-time AFISs efficiency and accuracy in the near future.

Systematic Mapping Study about entity reconciliation in Big Data.Eleven databases were consulted. Rigorous process where 2255 papers were analyzed leaving 61 primary studies.Analytic data statistics after the classification of the primary studies.Discussion that presents some interesting intelligent system-based papers for solving entity reconciliation and conclusions obtained from the data retrieved. The entity reconciliation (ER) problem aroused much interest as a research topic in today's Big Data era, full of big and open heterogeneous data sources. This problem poses when relevant information on a topic needs to be obtained using methods based on: (i) identifying records that represent the same real world entity, and (ii) identifying those records that are similar but do not correspond to the same real-world entity. ER is an operational intelligence process, whereby organizations can unify different and heterogeneous data sources in order to relate possible matches of non-obvious entities. Besides, the complexity that the heterogeneity of data sources involves, the large number of records and differences among languages, for instance, must be added. This paper describes a Systematic Mapping Study (SMS) of journal articles, conferences and workshops published from 2010 to 2017 to solve the problem described before, first trying to understand the state-of-the-art, and then identifying any gaps in current research. Eleven digital libraries were analyzed following a systematic, semiautomatic and rigorous process that has resulted in 61 primary studies. They represent a great variety of intelligent proposals that aim to solve ER. The conclusion obtained is that most of the research is based on the operational phase as opposed to the design phase, and most studies have been tested on real-world data sources, where a lot of them are heterogeneous, but just a few apply to industry. There is a clear trend in research techniques based on clustering/blocking and graphs, although the level of automation of the proposals is hardly ever mentioned in the research work.

An unsupervised multilingual approach to identify topics on Twitter is proposed.Localised language can be leveraged for identifying relevant and important topics.Joint term ranking coupled with DPMM clustering consistently performed well.Multilingual sentiment analysis is essential to understand sentiment on the ground.Topics coverage of social media and main stream media does not always stay the same. Social media data can be valuable in many ways. However, the vast amount of content shared and the linguistic variants of languages used on social media are making it very challenging for high-value topics to be identified. In this paper, we present an unsupervised multilingual approach for identifying highly relevant terms and topics from the mass of social media data. This approach combines term ranking, localised language analysis, unsupervised topic clustering and multilingual sentiment analysis to extract prominent topics through analysis of Twitter's tweets from a period of time. It is observed that each of the ranking methods tested has their strengths and weaknesses, and that our proposed Joint ranking method is able to take advantage of the strengths of the ranking methods. This Joint ranking method coupled with an unsupervised topic clustering model is shown to have the potential to discover topics of interest or concern to a local community. Practically, being able to do so may help decision makers to gauge the true opinions or concerns on the ground. Theoretically, the research is significant as it shows how an unsupervised online topic identification approach can be designed without much manual annotation effort, which may have great implications for future development of expert and intelligent systems.

Synthetic Aperture Radars (SAR) are the main instrument used to support oil detection systems. In the microwave spectrum, oil slicks are identified as dark spots, regions with low backscatter at sea surface. Automatic and semi-automatic systems were developed to minimize processing time, the occurrence of false alarms and the subjectivity of human interpretation. This study presents an intelligent hybrid system, which integrates automatic and semi-automatic procedures to detect dark spots, in six steps: (I) SAR pre-processing; (II) Image segmentation; (III) Feature extraction and selection; (IV) Automatic clustering analysis; (V) Decision rules and, if needed; (VI) Semi-automatic processing. The results proved that the feature selection is essential to improve the detection capability, keeping only five pattern features to automate the clustering procedure. The semi-automatic method gave back more accurate geometries. The automatic approach erred more including regions, increasing the dark spots area, while the semi-automatic method erred more excluding regions. For well-defined and contrasted dark spots, the performance of the automatic and the semi-automatic methods is equivalent. However, the fully automatic method did not provide acceptable geometries in all cases. For these cases, the intelligent hybrid system was validated, integrating the semi-automatic approach, using compact and simple decision rules to request human intervention when needed. This approach allows for the combining of benefits from each approach, ensuring the quality of the classification when fully automatic procedures are not satisfactory.

FCTFS works in both autonomous and user guided mode.The defined taxonomy helps in arriving at optimal number of good quality clusters.Feature elimination due to irrelevance and redundancy is clearly isolated.It is faster than traditional search based methods.Yields superior results compared to some state of the art methods over 24 data sets. Feature subset selection is basically an optimization problem for choosing the most important features from various alternatives in order to facilitate classification or mining problems. Though lots of algorithms have been developed so far, none is considered to be the best for all situations and researchers are still trying to come up with better solutions. In this work, a flexible and user-guided feature subset selection algorithm, named as FCTFS (Feature Cluster Taxonomy based Feature Selection) has been proposed for selecting suitable feature subset from a large feature set. The proposed algorithm falls under the genre of clustering based feature selection techniques in which features are initially clustered according to their intrinsic characteristics following the filter approach. In the second step the most suitable feature is selected from each cluster to form the final subset following a wrapper approach. The two stage hybrid process lowers the computational cost of subset selection, especially for large feature data sets. One of the main novelty of the proposed approach lies in the process of determining optimal number of feature clusters. Unlike currently available methods, which mostly employ a trial and error approach, the proposed method characterises and quantifies the feature clusters according to the quality of the features inside the clusters and defines a taxonomy of the feature clusters. The selection of individual features from a feature cluster can be done judiciously considering both the relevancy and redundancy according to users intention and requirement. The algorithm has been verified by simulation experiments with different bench mark data set containing features ranging from 10 to more than 800 and compared with other currently used feature selection algorithms. The simulation results prove the superiority of our proposal in terms of model performance, flexibility of use in practical problems and extendibility to large feature sets. Though the current proposal is verified in the domain of unsupervised classification, it can be easily used in case of supervised classification.