Spatiotemporal Kernel Density Visualization (STKDV) has been widely used across various domains in geospatial analysis, e.g., urban planning, traffic/traffic accident hotspot analysis, crime hotspot analysis, and disease spread modeling.However, STKDV is a computationally expensive tool, which has been complained by many domain experts.Although many recent solutions, including the sliding-window-based solution (SWS) and the prefix-matrix-based solution (PREFIX), have been proposed for improving the efficiency of generating an exact STKDV, these solutions still cannot be scalable to handle large-scale location datasets.To tackle this efficiency issue, we propose the pioneering block compression solution, called COMP, which can compress (or represent) a location dataset by a small amount of blocks.By combining COMP with the existing exact solutions, i.e., SWS and PREFIX, we show that COMP SWS and COMP PREFIX can generate approximate STKDV with an 𝜖-absolute error guarantee based on properly tuning the block size.Experimental results on four large-scale location datasets (up to 6.782 million data points) also verify that COMP SWS and COMP PREFIX can achieve speedups of 4.1x to 677.16x and 1.45x to 143.52x compared with SWS and PREFIX, respectively, without degrading the visualization results.The code of this paper can be found in https://github.com/YovelaZ/COMP.

The perception and generation of Human-Object Interaction (HOI) are crucial for fields such as robotics, AR/VR, and human behavior understanding. However, current approaches model this task in an offline setting, where information at each time step can be drawn from the entire interaction sequence. In contrast, in real-world scenarios, the information available at each time step comes only from the current moment and historical data, i.e., an online setting. We find that offline methods perform poorly in an online context. Based on this observation, we propose two new tasks: Online HOI Generation and Perception. To address this task, we introduce the OnlineHOI framework, a network architecture based on the Mamba framework that employs a memory mechanism. By leveraging Mamba's powerful modeling capabilities for streaming data and the Memory mechanism's efficient integration of historical information, we achieve state-of-the-art results on the Core4D and OAKINK2 online generation tasks, as well as the online HOI4D perception task.

Current big data clustering methods pose a challenge to efficiently estimate the number of clusters when dealing with a large number of instances and dimensions. However, determining the large number of clusters in a high-dimensional dataset is difficult due to the conflicting nature of its subspaces. In this paper, we propose a new distributed clustering ensemble method calledsubspaceclusterensemble (SSCE) to find the number of clusters in big high-dimensional datasets with informative subspace clusters. We represent the high-dimensional dataset as a set of random subspaces, and then in each subspace view, we investigate the number of clusters and initial cluster centers with the I-niceDP clustering scheme. To efficiently handle big data (i.e., a large number of instances with high dimensions), we use multiple random samples of a big dataset to estimate the number of clusters. With this aim, we first adopt the random sample partitioning (RSP) data model to facilitate random sample generation and then apply the subspace cluster estimation scheme. To do so, each random sample (i.e., RSP data block) is processed independently in parallel, and subspace clustering components are generated. A novel two-stage clustering ensemble, ball fusion (BF), is proposed to estimate the final result from the subspace clustering components and multiple RSP sample outcomes. Subspace and full-space clusters induce key connections within subspaces to generate the final ensemble outcome. Results from experiments on synthetic and real-world datasets demonstrated the effectiveness of the proposed method, which outperformed the state-of-the-art baselines.

Summary Clustering ensemble is a popular approach for identifying data clusters that combines the clustering results from multiple base clustering algorithms to produce more accurate and robust data clusters. However, the performance of clustering ensemble algorithms is highly dependent on the quality of clustering members. To address this problem, this paper proposes a member enhancement‐based clustering ensemble (MECE) algorithm that selects the ensemble members by considering their distribution consistency. MECE has two main components, called heterocluster splitting and homocluster merging. The first component estimates two probability density functions (p.d.f.s) estimated on the sample points of an heterocluster and represents them using a Gaussian distribution and a Gaussian mixture model. If the random numbers generated by these two p.d.f.s have different probability distributions, the heterocluster is then split into smaller clusters. The second component merges the clusters that have high neighborhood densities into a homocluster, where the neighborhood density is measured using a novel evaluation criterion. In addition, a co‐association matrix is presented, which serves as a summary for the ensemble of diverse clusters. A series of experiments were conducted to evaluate the feasibility and effectiveness of the proposed ensemble member generation algorithm. Results show that the proposed MECE algorithm can select high quality ensemble members and as a result yield the better clusterings than six state‐of‐the‐art ensemble clustering algorithms, that is, cluster‐based similarity partitioning algorithm (CSPA), meta‐clustering algorithm (MCLA), hybrid bipartite graph formulation (HBGF), evidence accumulation clustering (EAC), locally weighted evidence accumulation (LWEA), and locally weighted graph partition (LWGP). Specifically, MECE algorithm has the nearly 23% higher average NMI, 27% higher average ARI, 15% higher average FMI, and 10% higher average purity than CSPA, MCLA, HBGF, EAC, LWEA, and LWGA algorithms. The experimental results demonstrate that MECE algorithm is a valid approach to deal with the clustering ensemble problems.

Effective online portfolio selection necessitates seamless integration of three key properties: diversity, sparsity , and risk control. However, existing algorithms often prioritize one property at the expense of the others due to inherent conflicts. To address this issue, we propose an asset subset-constrained minimax (ASCM) optimization framework, which generates optimal portfolios from diverse investment strategies represented as asset subsets. ASCM consists of: (i) a minimax optimization model that focuses on risk control by considering a set of loss functions constrained by different asset subsets; (ii) the construction of asset subsets via price-feature clipping, which effectively reduces redundant assets in the portfolio; (iii) a state-based estimation of price trends that guides all ASCM loss functions, facilitating the generation of sparse solutions. We solve the ASCM minimax model using an efficient iterative updating formula derived from the projected subgradient method. Furthermore, we achieve near O ( 1 ) time complexity through a novel initialization scheme. Experimental results demonstrate that ASCM outperforms eight representative algorithms, including the best constant rebalanced portfolio in hindsight (BCRP) on five out of the six real-world financial datasets. Notably, ASCM achieves a 67-fold improvement over BCRP in cumulative wealth on the TSE dataset.

The sampling-based approximation method has demonstrated its potential in various domains such as machine learning, query processing, and data analysis. Most preceding sampling algorithms generate samples at the record level, making it impractical to apply them to very large datasets using a single machine. Even distributed solutions encounter efficiency issues when dealing with terabyte-scale datasets. In this paper, we introduce a scalable sampling approach named CDFRS, which can generate samples with a distribution-preserving guarantee from extensive datasets. CDFRS exhibits significantly improved speed compared to existing sampling algorithms when dealing with terabyte-scale datasets. We provide theoretical guarantees and empirical justifications, demonstrating that samples generated by the CDFRS approach maintain the distribution characteristics of the original dataset. Additionally, we propose a sample size determination algorithm, denoted as A2. Experiment results indicate that the running time of CDFRS shows at least an order of magnitude improvement over other distributed sampling methods. Notably, sampling a 10TB dataset using CDFRS only takes hundreds of seconds, while the compared method requires more than ten thousand seconds. In the context of big data analysis, including tasks such as classification and clustering, models trained with samples generated by CDFRS closely match those trained with the entire training set. Furthermore, the proposed A2 algorithm efficiently determines an appropriate sample size compared with traditional methods.

Basket analysis is a prevailing technique to help retailers uncover patterns and associations of sold products in customer shopping transactions. However, as the size of transaction databases grows, the traditional basket analysis techniques and systems become less effective because of two issues in the applications of the big data age: data scalability and flexibility to adapt different application tasks. This paper proposes a scalable distributed frequent itemset mining (ScaDistFIM) algorithm for basket analysis on big transaction data to solve these two problems. ScaDistFIM is performed in two stages. The first stage uses the FP-Growth algorithm to compute the local frequent itemsets from each random subset of the distributed transaction dataset, and all random subsets are computed in parallel. The second stage uses an approximation method to aggregate all local frequent itemsets to the final approximate set of frequent itemsets where the support values of the frequent itemsets are estimated. We further elaborate on implementing the ScaDistFIM algorithm and a flexible basket analysis system using Spark SQL queries to demonstrate the system’s flexibility in real applications. The experiment results on synthetic and real-world transaction datasets demonstrate that compared to the Spark FP-Growth algorithm, the ScaDistFIM algorithm can achieve time savings of at least 90% while ensuring nearly 100% accuracy. Hence, the ScaDistFIM algorithm exhibits superior scalability. On dataset GenD with 1 billion records, the ScaDistFIM algorithm requires only 360 s to achieve 100% precision and recall. In contrast, due to memory limitations, Spark FP-Growth cannot complete the computation task.

The synthetic minority oversampling technique (SMOTE) is a popular algorithm to reduce the impact of class imbalance in building classifiers, and has received several enhancements over the past 20 years. SMOTE and its variants synthesize a number of minority-class sample points in the original sample space to alleviate the adverse effects of class imbalance. This approach works well in many cases, but problems arise when synthetic sample points are generated in overlapping areas between different classes, which further complicates classifier training. To address this issue, this paper proposes a novel generalization-oriented rather than imputation-oriented minority-class sample point generation algorithm, named overlapping minimization SMOTE (OM-SMOTE). This algorithm is designed specifically for binary imbalanced classification problems. OM-SMOTE first maps the original sample points into a new sample space by balancing sample encoding and classifier generalization. Then, OM-SMOTE employs a set of sophisticated minority-class sample point imputation rules to generate synthetic sample points that are as far as possible from overlapping areas between classes. Extensive experiments have been conducted on 32 imbalanced datasets to validate the effectiveness of OM-SMOTE. Results show that using OM-SMOTE to generate synthetic minority-class sample points leads to better classifier training performances for the naive Bayes, support vector machine, decision tree, and logistic regression classifiers than the 11 state-of-the-art SMOTE-based imputation algorithms. This demonstrates that OM-SMOTE is a viable approach for supporting the training of high-quality classifiers for imbalanced classification. The implementation of OM-SMOTE is shared publicly on the GitHub platform at https://github.com/luxuan123123/OM-SMOTE/ .

Deep Forest, a powerful alternative to deep neural networks, has gained much attention due to its advantages, such as low complexity, minimal hyperparameter requirements, and strong application performance. In the current big data environment, where data volumes and model complexities are growing rapidly, distributed computing is needed to increase computational efficiency. Recently, a distributed deep forest approach, called BLB-gcForest (Bag of Little Bootstraps-gcForest), has been successful in reducing training instances within cascade forests, combining BLB and granularity segmentation, thus improving the computational efficiency and scalability of distributed deep forests. However, it still transmits the entire data set between layers and requires double sampling with BLB, limiting the amount of data and the scalability of resource utilization. This paper introduces a novel algorithm, RSP-gcForest, based on Random Sample Partition (RSP) to improve distributed deep forests computational efficiency and scalability. RSP-gcForest uses block-level samples that replace the full dataset, significantly reducing interlayer instance transmission and prediction within cascade forests. Additionally, RSP blocks are integrated with the segmentation granularity of cascade forests for ensemble learning, effectively addressing computational efficiency and resource constraints. We conducted experiments on four extensive datasets using Spark and evaluated performance across five key metrics. The results clearly show that RSP-gcForest, while maintaining high classification quality, surpasses state-of-the-art methods in terms of computational efficiency and resource utilization. Furthermore, it achieves superior load balancing, demonstrating its potential as a powerful tool in big data and distributed computing.

Dimensionality reduction through feature selection becomes inevitable to overcome the problem of the Curse of dimensionality. In this article, we propose a feature (gene) selection method for high dimensional gene expression (GE) data through a Multi-objective optimization-based Multi-view Co-Clustering algorithm (named MMCo-Clus). A thorough comparative analysis with existing feature selection algorithms using external/internal evaluation metrics supports our proposed method’s potency.

Clustering a big distributed dataset of hundred gigabytes or more is a challenging task in distributed computing. A popular method to tackle this problem is to use a random sample of the big dataset to compute an approximate result as an estimation of the true result computed from the entire dataset. In this paper, instead of using a single random sample, we use multiple random samples to compute an ensemble result as the estimation of the true result of the big dataset. We propose a distributed computing framework to compute the ensemble result. In this framework, a big dataset is represented in the RSP data model as random sample data blocks managed in a distributed file system. To compute the ensemble clustering result, a set of RSP data blocks is randomly selected as random samples and clustered independently in parallel on the nodes of a cluster to generate the component clustering results. The component results are transferred to the master node, which computes the ensemble result. Since the random samples are disjoint and traditional consensus functions cannot be used, we propose two new methods to integrate the component clustering results into the final ensemble result. The first method uses component cluster centers to build a graph and the METIS algorithm to cut the graph into subgraphs, from which a set of candidate cluster centers is found. A hierarchical clustering method is then used to generate the final set of$k$cluster centers. The second method uses the clustering-by-passing-messages method to generate the final set of$k$cluster centers. Finally, the$k$-means algorithm was used to allocate the entire dataset into$k$clusters. Experiments were conducted on both synthetic and real-world datasets. The results show that the new ensemble clustering methods performed better than the comparison methods and that the distributed computing framework is efficient and scalable in clustering big datasets.

Serverless Computing and Function-as-a-Service (FaaS) offer convenient and transparent services to developers and users. The deployment and resource allocation of services are managed by the cloud service providers. Meanwhile, the development of smart mobile devices and network technology enables the collection and transmission of a huge amount of data, which creates the mobile edge computing shifting tasks to the network edge for mobile users. In this paper, we propose a deviceless edge computing system targeting the mobility of end users. We focus on the migration of virtual functions to provide uninterrupted services to mobile users. We introduce the deviceless edge computing model and propose a seamless migration scheme of virtual functions with limited involvement of function developers. We formulate the migration decision problem into integer linear programming and use receding horizon control (RHC) for online solutions. We implement the migration system and algorithm to support delay-sensitive scenarios over real edge devices and develop a streaming game as the virtual function to test the performance. Extensive experiments in real scenarios exhibit the system has the ability to support high-mobility and delay-sensitive application scenarios. Extensive simulation results also show its applicability over large-scale networks.

Structured sequences are a popular data representation, used to model complex data such as traffic networks. A key machine learning task for structured sequences is node classification, that is predicting the class labels of unlabeled nodes. Though many node classification models were proposed, they assume a closed world setting, that all class labels appear in the training data. But in the real-world, the presence of never-before-seen class labels in testing data can considerably degrade a classifier’s accuracy. A promising solution to this issue is to build classifiers for an open-world setting, where samples with unknown class labels are continuously observed such that training and testing data may have different class label spaces. Several approaches have been proposed for open-world learning problems in computer vision and natural language processing, but they cannot be applied directly to structured sequences due to the complexity of their non-Euclidean properties and their dynamic nature. This paper addresses this important research gap by proposing a novel Open-world Structured Sequence node Classification (OSSC) model, to learn from structured sequences in an open-world setting. OSSC captures the structural and temporal information via a GCN-based dynamic variational framework. A latent distribution sequence is learned for each node using both stochastic states and deterministic states, to capture the evolution of node attributes and topology, followed by a sampling process to generate node representations. An open-world classification loss is further adopted to ensure that node representations are sensitive to unknown classes. And a combination of Openmax and Softmax is utilized to recognize nodes from unknown classes and to classify others to one of the known classes. Experiments on real-world datasets show that the proposed OSSC method is capable of learning accurate open-world node classifiers from structured sequence data.

To alleviate the attribute independence assumption, this paper proposes an auto-encoding naive Bayesian classifier (AE-NBC) which is trained based on transformed independent condition-attributes using an auto-encoding strategy. First, an auto-encoding neural network (AENN) with single hidden-layer is created to map the original data set to a high-dimensional data space. To guarantee the independence among auto-encoding condition-attributes (AECAs), an efficient and convergent objective function is elaborated to optimize the input-layer and output-layer weights of the AENN by considering both the auto-encoding error and attribute independence degree. Second, independent component analysis is iteratively carried out to remove the redundant attributes in AECAs so that the optimal transformed condition-attributes (TCAs) are obtained for the training of the NBC. Experiments were conducted on 29 benchmark data sets to validate the rationality, feasibility, and effectiveness of AE- NBC. Experimental results show that the independence among TCAs is stronger than among the original condition-attributes, and that AE- NBC can obtain higher training and testing accuracies than six Bayesian classifiers.

This article presents a Bayesian attributebagging-based extreme learning machine (BAB-ELM)to handle high-dimensional classification and regression problems. First, thedecision-making degree (DMD)of a condition attribute is calculated based on the Bayesian decision theory, i.e., the conditional probability of the condition attribute given the decision attribute. Second, the condition attribute with the highest DMD is put into thecondition attribute group (CAG)corresponding to the specific decision attribute. Third, thebagging attribute groups (BAGs)are used to train an ensemble learning model ofextreme learning machines (ELMs).Each base ELM is trained on a BAG which is composed of condition attributes that are randomly selected from the CAGs. Fourth, the information amount ratios of bagging condition attributes to all condition attributes is used as the weights to fuse the predictions of base ELMs in BAB-ELM. Exhaustive experiments have been conducted to compare the feasibility and effectiveness of BAB-ELM with seven other ELM models, i.e., ELM, ensemble-based ELM (EN-ELM), voting-based ELM (V-ELM), ensemble ELM (E-ELM), ensemble ELM based on multi-activation functions (MAF-EELM), bagging ELM, and simple ensemble ELM. Experimental results show that BAB-ELM is convergent with the increase of base ELMs and also can yield higher classification accuracy and lower regression error for high-dimensional classification and regression problems.

This article proposes a deep learning solution to the online portfolio selection problem based on learning a latent structure directly from a price time series. It introduces a novel wealth flow matrix for representing a latent structure that has special regular conditions to encode the knowledge about the relative strengths of assets in portfolios. Therefore, a wealth flow model (WFM) is proposed to learn wealth flow matrices and maximize portfolio wealth simultaneously. Compared with existing approaches, our work has several distinctive benefits: (1) the learning of wealth flow matrices makes our model more generalizable than models that only predict wealth proportion vectors, and (2) the exploitation of wealth flow matrices and the exploration of wealth growth are integrated into our deep reinforcement algorithm for the WFM. These benefits, in combination, lead to a highly-effective approach for generating reasonable investment behavior, including short-term trend following, the following of a few losers, no self-investment, and sparse portfolios. Extensive experiments on five benchmark datasets from real-world stock markets confirm the theoretical advantage of the WFM, which achieves the Pareto improvements in terms of multiple performance indicators and the steady growth of wealth over the state-of-the-art algorithms.

In the era of Big Data, cluster analysis of high-dimensional data sets often suffers from theCurse of dimensionality. To overcome this problem, the dimensionality reduction throughfeature selectionbecomes inevitable. Co-clustering or two-way clustering is considered to be a more sophisticated tool than conventional one-way clustering. Moreover, the advent of multi-view learning shows that the subjects of a data set can be interpreted in many ways. Interestingly, a minimal number of existing feature selection algorithms take advantage of the co-clustering method and are designed to consider multi-view data. Motivated by this, in the current article, we propose a feature (gene) selection method for high dimensional gene expression (GE) data through amulti-objective optimization basedmulti-viewCo-Clustering algorithm (namedMMCo-Clus). A popular evolutionary technique – Non-dominated Sorting Genetic Algorithm-II (NSGA-II) has been utilized as the proposed method's underlying optimization strategy. First, we construct two views of a chosen data set, utilizing knowledge from two different biological data sources. Next, we develop the MMCo-Clusalgorithm considering the constructed views to identify a set of “good” co-clustering solutions. Finally, based on a concept ofconsensus operationon the co-clustering outcome, a small number of most relevant and non-redundant features are extracted from the original feature-space. The reduced dimension formed by new feature-space causes to decrease the computational burden and noise level of original data. For experimental analysis, we have chosen three benchmark GE data sets. Our feature selection method's effectiveness is evaluated through sample-classification accuracy, accompanied by the cluster profile plot/Eisen plot/t-SNE plot, and biological/statistical significance test. A thorough comparative analysis with existing feature selection algorithms using external and internal evaluation metrics supports our proposed method's potency.

A novel random sample partition-based clustering ensemble (RSP-CE) algorithm is proposed in this paper to handle the big data clustering problems. There are three key components in RSP-CE algorithm, i.e., generating the base clustering results on RSP data blocks, harmonizing the based clustering results with maximum mean discrepancy (MMD) criterion, and refining the RSP clustering results. RSP data blocks have the consistent sample distributions with the whole big data and thus provide the possibility for using base clustering results on different data subsets to approximate the clustering result on whole big data. The experimental results in comparison with other 5 well-known clustering ensemble algorithms on 4 big data sets show that RSP-CE algorithm obtains the better normalized mutual information (NMI) values and Fowlkes-Mallows Index (FMI) values with the less training time consumptions and thus demonstrate that RSP-CE algorithm is a viable approach to deal with the big data clustering problems.

Due to the ubiquitous presence of missing values in real-world datasets, an imputation algorithm can recover the missing values and provide users with a complete dataset that utilizes all the available observed information. However, most of the imputation methods still have several limitations, including cannot restore the original distribution, handling various data missing patterns, and high missing rate dataset. In this poster, a novel neural network-based two-stage missing value imputation (abbreviated as TS-MVI) method is proposed to fill an incomplete condition attribute with the optimized attribute values for the supervised learning task. By initializing the missing values with random numbers, the imputation values are iteratively adjusted based on the new updating rule by minimizing both the autoencoder-oriented objective function and neural network-based classification error. The persuasive experiments show that TS-MVl method significantly outperforms current state-of-the-art imputation methods and thus demonstrate TS-MVI is a viable approach to deal with the missing value imputation problem.

We propose a sampling-based method, called RSP-Hist, to construct approximate equi-width histograms and help data scientists explore the probability distribution of big data on Hadoop clusters. In RSP- Hist, the Random Sample Partition (RSP) model is used to store a big data set as ready-to-use random sample data blocks, called RSP blocks, in the Hadoop Distributed File System (HDFS). An approximate histogram is computed by applying a sequential histogram algorithm in parallel to each block in a block-level sample of RSP blocks. Local histograms from individual RSP blocks are combined to produce an approximate histogram for the entire data. We tested RSP-Hist on four data sets using a small computing cluster. In this paper, we demonstrate the effect of the sampling rate and the number of buckets on the histogram accuracy and show that RSP-based approximate histograms are equivalent to the exact histograms computed from the entire data. RSP- Hist can avoid the data correlation issue in HDFS blocks and significantly reduce both computation and communication costs. It enables iterative and interactive exploration of big data sets on small computing clusters and can be used for multivariate data exploration.

This paper proposes a compressed hidden naive Bayesian (C-HNB) classifier which is an improved version of hidden naive Bayesian (HNB) by compressing the Bayesian network structure and calculating the attribute correlation with maximal information coefficient (MIC). In C-HNB, we remodel the Bayesian network structure based on the chain rule of joint probability distribution so that the number of hidden parent nodes in C-HNB is smaller than the number of hidden parent nodes in HNB, which reduces the training complexity of Bayesian network. In addition, the attribute correlation in C-HNB is calculated with MIC rather than the mutual information, which makes the Bayesian network more stable because the calculation of mutual information is severely influenced by the discretization of continuous attribute. On the selected KEEL benchmark data sets, we compare the classification performances of C-HNB with HNB. The comparative results show that C-HNB can obtain the better prediction accuracy with the less training time in comparison with HNB and thus the experimental results demonstrate the effectiveness of C-HNB.

TripletGAN is a variant of Generative Adversarial Network (GAN) by replacing the classification loss of discriminator with a triplet loss. Although TripletGAN delivers better mode coverage than vanilla GAN thanks to the characteristics of adversarial triplet loss that maximizes the embedding distance between generated samples, its adversarial training method suffers from the drawback that some generated images tend to deviate from the real sample distribution and noisy images are produced as we increase the number of iterations of training. In this paper, we propose an adversarially balanced triplet loss with four dynamic coefficients to achieve a trade-off between the quality and the diversity of generated samples. We also design a novel network architecture to provide GANs with an auto-encoding ability. Extensive experiments demonstrate the effectiveness of our proposed methods in terms of alleviating the problem in TripletGAN and the superiority in terms of reconstruction over some methods that directly train generator and encoder such as O-GAN.

Classification problems in which the number of features (dimensions) is unduly higher than the number of samples (observations) is an essential research and application area in a variety of domains, especially in computational biology. It is also known as a high-dimensional small-sample-size (HDSSS) problem. Various dimensionality reduction methods have been developed, but they are not potent with the small-sample-sized high-dimensional datasets and suffer from overfitting and high-variance gradients. To overcome the pitfalls of sample size and dimensionality, this study employed variational autoencoder (VAE), which is a dynamic framework for unsupervised learning in recent years. The objective of this study is to investigate a reliable classification model for high-dimensional and small-sample-sized datasets with minimal error. Moreover, it evaluated the strength of different architectures of VAE on the HDSSS datasets. In the experiment, six genomic microarray datasets from Kent Ridge Biomedical Dataset Repository were selected, and several choices of dimensions (features) were applied for data preprocessing. Also, to evaluate the classification accuracy and to find a stable and suitable classifier, nine state-of-the-art classifiers that have been successful for classification tasks in high-dimensional data settings were selected. The experimental results demonstrate that the VAE can provide superior performance compared to traditional methods such as PCA, fastICA, FA, NMF, and LDA in terms of accuracy and AUROC.

One of the most significant problems in cluster analysis is to determine the number of clusters in unlabeled data, which is the input for most clustering algorithms. Some methods have been developed to address this problem. However, little attention has been paid on algorithms that are insensitive to the initialization of cluster centers and utilize variable weights to recover the number of clusters. To fill this gap, we extend the standard fuzzy k-means clustering algorithm. It can automatically determine the number of clusters by iteratively calculating the weights of all variables and the membership value of each object in all clusters. Two new steps are added to the fuzzy k-means clustering process. One of them is to introduce a penalty term to make the clustering process insensitive to the initial cluster centers. The other one is to utilize a formula for iterative updating of variable weights in each cluster based on the current partition of data. Experimental results on real-world and synthetic datasets have shown that the proposed algorithm effectively determined the correct number of clusters while initializing the different number of cluster centroids. We also tested the proposed algorithm on gene data to determine a subset of important genes.

Although many spectral clustering algorithms have been proposed during the past decades, they are not scalable to large-scale data due to their high computational complexities. In this paper, we propose a novel spectral clustering method for large-scale data, namely, large-scale balanced min cut (LABIN). A new model is proposed to extend the self-balanced min-cut (SBMC) model with the anchor-based strategy and a fast spectral rotation with linear time complexity is proposed to solve the new model. Extensive experimental results show the superior performance of our proposed method in comparison with the state-of-the-art methods including SBMC.

To break the in-memory bottleneck and facilitate online sampling in cluster computing frameworks, we propose a new sampling-based system for approximate big data analysis on computing clusters. We address both computational and statistical aspects of big data across the main layers of cluster computing frameworks: big data storage, big data management, big data online sampling, big data processing, and big data exploration and analysis. We use the new Random Sample Partition (RSP) distributed data model to store a big data set as a set of ready-to-use random sample data blocks in Hadoop Distributed File System (HDFS), called RSP blocks. With this system, only a few RSP blocks are selected and processed using a sequential algorithm in a distributed data-parallel manner to produce approximate results for the entire data set. In this paper, we present a prototype RSP-based system and demonstrate its advantages. Our experiments show that RSP blocks can be used to get approximate models and summary statistics as well as estimate the proportions of inconsistent values without computing the entire data or running expensive online sampling operations. This new system enables big data exploration and analysis where the entire data set cannot be computed.

Underwater channels are considered challenging media in communication due to the harsh nature of such environments. However, dynamic transmission can assist in finding sub-optimal solutions by adaptively changing the employed techniques, e.g. the forwarding scheme between nodes and the transmitted signal intensity control, to compromise for the instantaneous fluctuations in various underwater environments. Additionally, Machine Learning (ML) techniques can provide appropriate solutions for various problems e.g. routing, resource allocation, and energy-efficiency to further enhance the quality of the communication systems. In this paper, we propose a novel dynamical transmission framework for multi-hop Internet of Underwater Things (IoUT) and underwater networks to fit for various conditions. The proposed framework employs a heuristic forwarding scheme selection approach beside an adaptive transmission signal intensity method. We also propose a decision-tree based ML-model that adaptively learns the proper forwarding method beside the appropriate amount of the transmitted signal intensity for each relay node to minimize the transmission error rate and the power consumption depending on numerous parameters e.g. node location, link reliability and certain water quality metrics such as water temperature, depth, and pH measurements. The model achieves remarkable accuracy for training and testing patterns beyond the 99%.

With wide adoption of Internet of Things (IoT) across the world, the IoT devices are facing more and more intensive computation task nowadays. However, the IoT devices are usually limited by their computing capability and battery lifetime. Mobile edge computing provides new opportunities for developments of IoT, since edge computing servers which are close to devices can provide more powerful computing resources. The IoT devices can offload the intensive computing tasks to edge computing servers, while saving their own computing resources and reducing energy consumption. However, the benefits come at the cost of higher latency, mainly due to additional transmission time, and it may be unacceptable for many IoT applications. In this paper, we try to find a tradeoff between the energy consumption and latency, in order to satisfy user demands of various IoT applications. We formalize the problem into a constrained multiobjective optimization problem and find the optimal solutions by a modified fast and elitist nondominated sorting genetic algorithm (NSGA-II). To improve the performance of the algorithm, we propose a novel problem-specific encoding scheme and genetic operators in the proposed modified NSGA-II. We also conduct extensive simulation experiments to evaluate the proposed algorithm and its sensitivity under certain major parameters. The experimental results show that the proposed algorithm can find a large number of optimal solutions to adjust the corresponding offloading decision according to the real-world situation.

Microarray technology enables the collection of vast amounts of gene expression data from biological experiments. Clustering algorithms have been successfully applied to exploring the gene expression data. Since a set of genes may be only correlated to a subset of samples, it is useful to use co-clustering to recover co-clusters in the gene expression data. In this paper, we propose a novel algorithm, called Subspace Weighting Co-Clustering (SWCC), for high dimensional gene expression data. In SWCC, a gene subspace weight matrix is introduced to identify the contribution of gene objects in distinguishing different sample clusters. We design a new co-clustering objective function to recover the co-clusters in the gene expression data, in which the subspace weight matrix is introduced. An iterative algorithm is developed to solve the objective function, in which the subspace weight matrix is automatically computed during the iterative co-clustering process. Our empirical study shows encouraging results of the proposed algorithm in comparison with six state-of-the-art clustering algorithms on ten gene expression data sets. We also propose to use SWCC for gene clustering and selection. The experimental results show that the selected genes can improve the classification performance of Random Forests.

With the ever-increasing volume of data, alternative strategies are required to divide big data into statistically consistent data blocks that can be used directly as representative samples of the entire data set in big data analysis. In this paper, we propose the Random Sample Partition (RSP) distributed data model to represent a big data set as a set of disjoint data blocks, called RSP blocks. Each RSP block has a probability distribution similar to that of the entire data set. RSP blocks can be used to estimate the statistical properties of the data and build predictive models without computing the entire data set. We demonstrate the implications of the RSP model on sampling from big data and introduce a new RSP-based method for approximate big data analysis which can be applied to different scenarios in the industry. This method significantly reduces the computational burden of big data and increases the productivity of data scientists.

During the past decades, many spectral clustering algorithms have been proposed. However, their high computational complexities hinder their applications on large-scale data. Moreover, most of them use a two-step approach to obtain the optimal solution, which may deviate from the solution by directly solving the original problem. In this paper, we propose a new optimization algorithm, namely Direct Normalized Cut (DNC), to directly optimize the normalized cut model. DNC has a quadratic time complexity, which is a significant reduction comparing with the cubic time complexity of the traditional spectral clustering. To cope with large-scale data, a Fast Normalized Cut (FNC) method with linear time and space complexities is proposed by extending DNC with an anchor-based strategy. In the new method, we first seek a set of anchors and then construct a representative similarity matrix by computing distances between the anchors and the whole data set. To find high quality anchors that best represent the whole data set, we propose a Balanced k-means (BKM) to partition a data set into balanced clusters and use the cluster centers as anchors. Then DNC is used to obtain the final clustering result from the representative similarity matrix. A series of experiments were conducted on both synthetic data and real-world data sets, and the experimental results show the superior performance of BKM, DNC and FNC.

In this paper, we propose a personalized dialogue generation system, which combines reinforcement learning techniques with an attention-based hierarchical recurrent encoderdecoder model. Firstly, we incorporate user-specific information into the decoder to capture user's background information and speaking style. Secondly, we employ reinforcement learning techniques to maximize future reward in dialogue, which enables our system to generate topic-coherent, informative and grammatical responses. Moreover, we propose three types of rewards to characterize good conversations. Finally, we compare the performance of the following reinforcement learning methods in dialogue generation: policy gradient, Q-learning, and actor-critic algorithms. We conduct experiments to verify the effectiveness of the proposed model on two dialogue datasets. Experimental results demonstrate that our model can generate better personalized dialogues for different users. Quantitatively, our method achieves better performance than the state-of-the-art dialogue systems in terms of BLEU score, perplexity, and human evaluation.

Clustering of customer transaction data is an important procedure to analyze customer behaviors in retail and e-commerce companies. Note that products from companies are often organized as a product tree, in which the leaf nodes are goods to sell, and the internal nodes (except root node) could be multiple product categories. Based on this tree, we propose the “personalized product tree”, named purchase tree, to represent a customer's transaction records. So the customers' transaction data set can be compressed into a set of purchase trees. We propose a partitional clustering algorithm, named PurTreeClust, for fast clustering of purchase trees. A new distance metric is proposed to effectively compute the distance between two purchase trees. To cluster the purchase tree data, we first rank the purchase trees as candidate representative trees with a novel separate density, and then select the top k customers as the representatives of k customer groups. Finally, the clustering results are obtained by assigning each customer to the nearest representative. We also propose a gap statistic based method to evaluate the number of clusters. A series of experiments were conducted on ten real-life transaction data sets, and experimental results show the superior performance of the proposed method.

In data mining, objects are often represented by a set of features, where each feature of an object has only one value. However, in reality, some features can take on multiple values, for instance, a person with several job titles, hobbies, and email addresses. These features can be referred to as set-valued features and are often treated with dummy features when using existing data mining algorithms to analyze data with set-valued features. In this paper, we propose an SV- $k$ -modes algorithm that clusters categorical data with set-valued features. In this algorithm, a distance function is defined between two objects with set-valued features, and a set-valued mode representation of cluster centers is proposed. We develop a heuristic method to update cluster centers in the iterative clustering process and an initialization algorithm to select the initial cluster centers. The convergence and complexity of the SV- $k$ -modes algorithm are analyzed. Experiments are conducted on both synthetic data and real data from five different applications. The experimental results have shown that the SV- $k$ -modes algorithm performs better when clustering real data than do three other categorical clustering algorithms and that the algorithm is scalable to large data.

Most feature selection methods first compute a similarity matrix by assigning a fixed value to pairs of objects in the whole data or to pairs of objects in a class or by computing the similarity between two objects from the original data. The similarity matrix is fixed as a constant in the subsequent feature selection process. However, the similarities computed from the original data may be unreliable, because they are affected by noise features. Moreover, the local structure within classes cannot be recovered if the similarities between the pairs of objects in a class are equal. In this paper, we propose a novel local adaptive projection (LAP) framework. Instead of computing fixed similarities before performing feature selection, LAP simultaneously learns an adaptive similarity matrix and a projection matrix with an iterative method. In each iteration, is computed from the projected distance with the learned and W is computed with the learned . Therefore, LAP can learn better projection matrix by weakening the effect of noise features with the adaptive similarity matrix. A supervised feature selection with LAP (SLAP) method and an unsupervised feature selection with LAP (ULAP) method are proposed. Experimental results on eight data sets show the superiority of SLAP compared with seven supervised feature selection methods and the superiority of ULAP compared with five unsupervised feature selection methods.

Many spectral clustering algorithms have been proposed and successfully applied to image data analysis such as content based image retrieval, image annotation, and image indexing. Conventional spectral clustering algorithms usually involve a two-stage process: eigendecomposition of similarity matrix and clustering assignments from eigenvectors by k-means or spectral rotation. However, the final clustering assignments obtained by the two-stage process may deviate from the assignments by directly optimize the original objective function. Moreover, most of these methods usually have very high computational complexities. In this paper, we propose a new min-cut algorithm for image clustering, which scales linearly to the data size. In the new method, a self-balanced min-cut model is proposed in which the Exclusive Lasso is implicitly introduced as a balance regularizer in order to produce balanced partition. We propose an iterative algorithm to solve the new model, which has a time complexity of O(n) where n is the number of samples. Theoretical analysis reveals that the new method can simultaneously minimize the graph cut and balance the partition across all clusters. A series of experiments were conducted on both synthetic and benchmark data sets and the experimental results show the superior performance of the new method.

Many spectral clustering algorithms have been proposed and successfully applied to many high-dimensional applications. However, there are still two problems that need to be solved: 1) existing methods for obtaining the final clustering assignments may deviate from the true discrete solution, and 2) most of these methods usually have very high computational complexity. In this paper, we propose a Scalable Normalized Cut method for clustering of large scale data. In the new method, an efficient method is used to construct a small representation matrix and then clustering is performed on the representation matrix. In the clustering process, an improved spectral rotation method is proposed to obtain the solution of the final clustering assignments. A series of experimental were conducted on 14 benchmark data sets and the experimental results show the superior performance of the new method.

With the rapid increase of complex and high-dimensional sparse data, demands for new methods to select features by exploiting both labeled and unlabeled data have increased. Least regression based feature selection methods usually learn a projection matrix and evaluate the importances of features using the projection matrix, which is lack of theoretical explanation. Moreover, these methods cannot find both global and sparse solution of the projection matrix. In this paper, we propose a novel semi-supervised feature selection method which can learn both global and sparse solution of the projection matrix. The new method extends the least square regression model by rescaling the regression coefficients in the least square regression with a set of scale factors, which are used for ranking the features. It has shown that the new model can learn global and sparse solution. Moreover, the introduction of scale factors provides a theoretical explanation for why we can use the projection matrix to rank the features. A simple yet effective algorithm with proved convergence is proposed to optimize the new model. Experimental results on eight real-life data sets show the superiority of the method.

In many application areas, data that is being generated and processed goes beyond the petabyte scale. Analyzing such an increasing massive volume of data faces computational, as well as, statistical challenges. In order to solve these challenges, distributed and parallel processing frameworks have been used for implementing scalable data analysis algorithms. Nevertheless, processing the whole big data set at one time may exceed the available computing resources and the time requirements for some applications. Thus, approximate approaches can be used to achieve asymptotic analysis results, especially when data analysis algorithms are amenable to an approximate result rather than an exact one. However, most approximation approaches require taking a random sample of the data which is a nontrivial task when working with big data sets. In this paper, we employ ensemble learning as an approach for asymptotic analysis using randomly selected subsets (i.e. data blocks) of a big data set. We propose an asymptotic ensemble learning framework which depends on block-based sampling rather than record-based sampling. In order to demonstrate the feasibility and performance of this framework, we present an empirical analysis on real data sets. In addition to the scalability advantage, the experimental results show that several blocks of a data set are enough to get approximately the same results as those from using the whole data set.

Clustering of customer transaction data is usually an important procedure to analyze customer behaviors in retail and e-commerce companies. Note that products from companies are often organized as a product tree, in which the leaf nodes are goods to sell, and the internal nodes (except root node) could be multiple product categories. Based on this tree, we present to use a “personalized product tree”, called purchase tree, to represent a customer's transaction data. The customer transaction data set can be represented as a set of purchase trees. We propose a PurTreeClust algorithm for clustering of large-scale customers from purchase trees. We define a new distance metric to effectively compute the distance between two purchase trees from the entire levels in the tree. A cover tree is then built for indexing the purchase tree data and we propose a leveled density estimation method for selecting initial cluster centers from a cover tree. PurTreeClust, a fast clustering method for clustering of large-scale purchase trees, is then presented. Last, we propose a gap statistic based method for estimating the number of clusters from the purchase tree clustering results. A series of experiments were conducted on ten large-scale transaction data sets which contain up to four million transaction records, and experimental results have verified the effectiveness and efficiency of the proposed method. We also compared our method with three clustering algorithms, e.g., spectral clustering, hierarchical agglomerative clustering and DBSCAN. The experimental results have demonstrated the superior performance of the proposed method.

Clustering categorical sequences is an important and difficult data mining task. Despite recent efforts, the challenge remains, due to the lack of an inherently meaningful measure of pairwise similarity. In this paper, we propose a novel variable-order Markov framework, named weighted conditional probability distribution (WCPD), to model clusters of categorical sequences. We propose an efficient and effective approach to solve the challenging problem of model initialization. To initialize the WCPD model, we propose to use a first-order Markov model built on a weighted fuzzy indicator vector representation of categorical sequences, which we call the WFI Markov model. Based on a cascade optimization framework that combines the WCPD and WFI models, we design a new divisive hierarchical clustering algorithm for clustering categorical sequences. Experimental results on data sets from three different domains demonstrate the promising performance of our models and clustering algorithm.

Linear classification is an important technique in machine learning and data mining, and development of fast optimization methods for training linear classification models is a hot research topic. Stochastic gradient descent (SGD) can achieve relatively good results quickly, but unstable to converge. Limited-memory BFGS (L-BFGS) method converges, but takes a long time to train the model, as it needs to compute the gradient from the entire data set to make an update. In this paper, we investigate a hybrid method that integrates SGD and L-BFGS into a new optimization process SGD-LBFGS to take advantages of both optimization methods. In SGD-LBFGS, SGD is used to run initial iterations to obtain a suboptimal result, and then L-BFGS takes over to continue the optimization process until the process converges and a better model is built. We present a theoretical result to prove that SGD-LBFGS converges faster than SGD and L-BFGS. Experiment analysis on 6 real world data sets have shown that SGD-LBFGS converged 77% faster than L-BFGS on average and demonstrated more stable results than SGD.

This paper proposes TW-k-means, an automated two-level variable weighting clustering algorithm for multiview data, which can simultaneously compute weights for views and individual variables. In this algorithm, a view weight is assigned to each view to identify the compactness of the view and a variable weight is also assigned to each variable in the view to identify the importance of the variable. Both view weights and variable weights are used in the distance function to determine the clusters of objects. In the new algorithm, two additional steps are added to the iterative k-means clustering process to automatically compute the view weights and the variable weights. We used two real-life data sets to investigate the properties of two types of weights in TW-k-means and investigated the difference between the weights of TW-k-means and the weights of the individual variable weighting method. The experiments have revealed the convergence property of the view weights in TW-k-means. We compared TW-k-means with five clustering algorithms on three real-life data sets and the results have shown that the TW-k-means algorithm significantly outperformed the other five clustering algorithms in four evaluation indices.

In web mining, there are situations in which only few data is labeled which imposes difficulties on traditional web page classification algorithms. Active learning scheme is then proposed to sample the most representative unlabeled data, which are then annotated by external oracles. Most present active methods are based on series-mode query strategy, which deduces the process of active learning inefficient and unstable. In this paper, we propose a novel text oriented active semi-supervised classification model, which is so-called active SSC. Comparing with other active approaches, our model has the characteristic of comprehensibility, and thus it is easy to design a batch-mode query strategy. Experimental results on public text data showed our method is an effect and stable active approach.

In this paper, a new classification method (ADCC) for high-dimensional data is proposed. In this method, a decision cluster classification (DCC) model consists of a set of disjoint decision clusters, each labeled with a dominant class that determines the class of new objects falling in the cluster. A cluster tree is first generated from a training data set by recursively calling a variable weighting k-means algorithm. Then, the DCC model is extracted from the tree. Various tests including Anderson–Darling test are used to determine the stopping condition of the tree growing. A series of experiments on both synthetic and real data sets have been conducted. Their results show that the new classification method (ADCC) performed better in accuracy and scalability than existing methods like k-NN, decision tree and SVM. ADCC is particularly suitable for large, high-dimensional data with many classes.

Clustering categorical sequences remains an open and challenging task due to the lack of an inherently meaningful measure of pair wise similarity between sequences. Model initialization is an unsolved problem in model-based clustering algorithms for categorical sequences. In this paper, we propose a simple and effective Markov model to approximate the conditional probability distribution (CPD) model, and use it to design a novel two-tier Markov model to represent a sequence cluster. Furthermore, we design a novel divisive hierarchical algorithm for clustering categorical sequences based on the two-tier Markov model. The experimental results on the data sets from three different domains demonstrate the promising performance of our models and clustering algorithm.

This paper presents a changing window approach to exploring gene expression patterns in "snapshot windows". A snapshot window is a sub-matrix of co-expressed microarray data representing certain expression pattern. In this approach, we use a feature weighting k-means subspace clustering algorithm to generate a set of clusters and each cluster defines a set of "snapshot windows" which are characterized by different sets of ordered sample weights that were assigned by the clustering algorithm. We define an accumulated weighting threshold (AWT) as the sum of weights of samples in the "snapshot window". Given a cluster, different "snapshot windows" can be obtained by changing AWT to explore all possible local expression patterns in the cluster. Experiment results have shown our approach is effective and flexible in exploring various expression patterns and identifying novel ones.

This paper presents a comparison study of the fuzzy k-means algorithm and a new variant with variable weighting in clustering high dimensional data. The fuzzy k-means algorithm is effective in discovering the clusters with overlapping boundaries. However, this effectiveness can be handicapped in high dimensional data. The recent development of the k-means algorithm with automated variable weighting offers a new technique for dealing with high dimensional data that occurs in many new applications such as text mining and bioinformatics. In this paper, the variable weighting mechanism is incorporated in the fuzzy k-means algorithm to cluster high dimensional data with overlapping clusters. Experiments on real data sets have shown that the variable weighting fuzzy k-means produced better clustering results than the fuzzy k-means without variable weighting.

In this paper, we present an agglomerative fuzzy $k$-means clustering algorithm for numerical data, an extension to the standard fuzzy $k$-means algorithm by introducing a penalty term to the objective function to make the clustering process not sensitive to the initial cluster centers. The new algorithm can produce more consistent clustering results from different sets of initial clusters centers. Combined with cluster validation techniques, the new algorithm can determine the number of clusters in a data set, which is a well known problem in $k$-means clustering. Experimental results on synthetic data sets (2 to 5 dimensions, 500 to 5000 objects and 3 to 7 clusters), the BIRCH two-dimensional data set of 20000 objects and 100 clusters, and the WINE data set of 178 objects, 17 dimensions and 3 clusters from UCI, have demonstrated the effectiveness of the new algorithm in producing consistent clustering results and determining the correct number of clusters in different data sets, some with overlapping inherent clusters.

This paper presents a learning classifier system ensemble for knowledge discovery from incremental data. The new ensemble was designed with a two-level architecture to improve the generalization ability. The new incoming cases are first bootstrapped to generate data as inputs to the first level classical learning classifier systems. The second level contains a plurality-vote module to determine the final classification by combining the classification results of the first level learning classifier systems. Each learning classifier system in the first level consists of two major modules, a genetic algorithm module for facilitating rule-discovery and a reinforcement learning module for adjusting the strength of the corresponding rules when rewards are received from the environment. We propose a revised Wilson's compact rule algorithm for generation of the compact rule set from the population set to improve the readability of the model. Two experiments were conducted. One was data mining of medical data and the other was steganalysis of images. The experimental results have shown that the new ensemble produced better performance on incremental data mining and better generalization than the single learning classifier system and other supervised learning methods. The results also showed that the compact rules were more interpretable.

This correspondence describes extensions to the k-modes algorithm for clustering categorical data. By modifying a simple matching dissimilarity measure for categorical objects, a heuristic approach was developed in [4], [12] which allows the use of the k-modes paradigm to obtain a cluster with strong intrasimilarity and to efficiently cluster large categorical data sets. The main aim of this paper is to rigorously derive the updating formula of the k-modes clustering algorithm with the new dissimilarity measure and the convergence of the algorithm under the optimization framework.

This paper presents a new k-means type algorithm for clustering high-dimensional objects in sub-spaces. In high-dimensional data, clusters of objects often exist in subspaces rather than in the entire space. For example, in text clustering, clusters of documents of different topics are categorized by different subsets of terms or keywords. The keywords for one cluster may not occur in the documents of other clusters. This is a data sparsity problem faced in clustering high-dimensional data. In the new algorithm, we extend the k-means clustering process to calculate a weight for each dimension in each cluster and use the weight values to identify the subsets of important dimensions that categorize different clusters. This is achieved by including the weight entropy in the objective function that is minimized in the k-means clustering process. An additional step is added to the k-means clustering process to automatically compute the weights of all dimensions in each cluster. The experiments on both synthetic and real data have shown that the new algorithm can generate better clustering results than other subspace clustering algorithms. The new algorithm is also scalable to large data sets.

This paper proposes a new mediate module ADDI to organize the UDDI nodes published by the different corporations on supply chain and establish a UDDI dynamic ranking model. We suggest a novel model for supply chain conduct classification, and define the supply chain entity conduct types as a standard for Web service classification and UDDI ranking. In this model, Web service-oriented technologies and protocols are deployed for modeling, managing and executing business-oriented functionalities and environments. We focus on the efficient integration of supply chain as key points to harmonize these technologies. Agent-orientation concepts and technologies are applied for SCM construction and interaction patterns.

This paper proposes a k-means type clustering algorithm that can automatically calculate variable weights. A new step is introduced to the k-means clustering process to iteratively update variable weights based on the current partition of data and a formula for weight calculation is proposed. The convergency theorem of the new clustering process is given. The variable weights produced by the algorithm measure the importance of variables in clustering and can be used in variable selection in data mining applications where large and complex real data are often involved. Experimental results on both synthetic and real data have shown that the new algorithm outperformed the standard k-means type algorithms in recovering clusters in data.

Recently, Web services have become a new technology trend for Enterprise Application Integration (EAI) and more and more applications based on Web services are emerging. One of the problems in using Web services in business applications such as logistics is services composition automatically and efficiently. In this paper, we present a Dynamic, Demand-Driven Web services Engine called D3D-Serv to implement composite service functionality that is used to dynamically build composite services from existing services according to different business logics and requirements. In this D3D-Serv framework, the most challenging function to implement is dynamic selection of service providers at run time. The highly dynamic and distributed nature of Web services often makes some service providers overloaded at certain times while others idle. To solve this problem, we propose an efficient services selection and execution strategy that is based on the queuing theory and can provide guarantees for the QOS (Quality of Service) under provider's limited resources. Preliminary experimental results have shown that this algorithm is effective.

In this paper, we present an adaptive modelling technique for studying past behaviors of objects and predicting the near future events. Our approach is to define a sliding window (of different window sizes) over a time sequence and build autoregression models from subsequences in different windows. The models are representations of past behaviors of the sequence objects. We can use the AR coefficients as features to index subsequences to facilitate the query of subsequences with similar behaviors. We can use a clustering algorithm to group time sequences on their similarity in the feature space. We can also use the AR models for prediction within different windows. Our experiments show that the adaptive model can give better prediction than non-adaptive models.

This paper presents an empirical study on the visual method for cluster validation based on the Fastmap projection. The visual cluster validation method attempts to tackle two clustering problems in data mining: to verify partitions of data created by a clustering algorithm; and to identify genuine clusters from data partitions. They are achieved through projecting objects and clusters by Fastmap to the 2D space and visually examining the results by humans. A Monte Carlo evaluation of the visual method was conducted. The validation results of the visual method were compared with the results of two internal statistical cluster validation indices, which shows that the visual method is in consistence with the statistical validation methods. This indicates that the visual cluster validation method is indeed effective and applicable to data mining applications.

This correspondence describes extensions to the fuzzy k-means algorithm for clustering categorical data. By using a simple matching dissimilarity measure for categorical objects and modes instead of means for clusters, a new approach is developed, which allows the use of the k-means paradigm to efficiently cluster large categorical data sets. A fuzzy k-modes algorithm is presented and the effectiveness of the algorithm is demonstrated with experimental results.