Unsupervised Data Classification Research Articles

Configuration-Specific Insight into Single-Molecule Conductance and Noise Data Revealed by the Principal Component Projection Method.

We explore the merits of neural network boosted, principal-component-projection-based, unsupervised data classification in single-molecule break junction measurements, demonstrating that this method identifies highly relevant trace classes according to the well-defined and well-visualized internal correlations of the data set. To this end, we investigate single-molecule structures exhibiting double molecular configurations, exploring the role of the leading principal components in the identification of alternative junction evolution trajectories. We show how the proper principal component projections can be applied to separately analyze the high- or low-conductance molecular configurations, which we exploit in 1/f-type noise measurements on bipyridine molecules. This approach untangles the unclear noise evolution of the entire data set, identifying the coupling of the aromatic ring to the electrodes through the π orbitals in two distinct conductance regions, and its subsequent uncoupling as these configurations are stretched.

CSEA: A Fine-Grained Framework of Climate-Season-Based Energy-Aware in Cloud Storage Systems

Abstract Continuous data scale growth increases energy consumption and operating cost that cannot be ignored in cloud storage systems. Previous studies have shown that analyzing the characteristics of I/O access and mining data features is effective for reasonable data distribution in storage systems. The granularity and criterion of classification are the key factors in determining the data distribution. To decrease energy consumption and operating cost, this paper puts forward a fine-grained framework of the climatic-season-based energy-aware in cloud storage system called CSEA. The framework concludes the following three aspects: (i) data feature mining. CSEA discovers potential data features by analyzing data access to provide help with data classification. (ii) K-means clustering algorithm. CSEA uses an unsupervised data classification algorithm in machine learning to divide data into categories based on seasonal characteristics by gathering real I/O access. (iii) data distribution of fine-grained. On the basis of seasonal features, CSEA fuses regional features to further refine the data distribution granularity to save on energy consumption and operating cost. Simulation experiments using extended CloudSimDisk and the constructed mathematical models indicate that CSEA reduces the energy consumption and operating cost compared with the single data classification standard and coarse-grained data distribution.

Optimization of Human Resource Performance Management System Based on Improved R‐Means Clustering Algorithm

With the rapid development of network technology and database technology, computers have been able to store large‐scale and massive data. On the other hand, traditional data analysis and processing tools such as management information system can only process these data on the surface, but the deeper data analysis ability is not satisfactory. The contradiction between data supply ability and data analysis ability is becoming more and more prominent, so there is an urgent need for an automation technology that can deeply process data. Data mining technology came into being. Cluster analysis, as an important topic in data mining, is a data mining method that divides data into natural groups and gives the description of the characteristics of each group. It is a basic method of data mining and knowledge discovery. Cluster analysis is a data mining technology for unsupervised classification of data without prior knowledge and guidance. Through the appropriate use of advanced algorithms, it can explore the hidden valuable information, improve the quality of data analysis and interpretation, and provide a scientific judgment basis for the reprocessing or understanding of data by other data analysis and sorting tools. First, this paper briefly introduces the principle, development, and methods of cluster analysis and expounds the application of cluster analysis. Then it expounds the principle of R‐means clustering algorithm, analyzes the advantages and disadvantages of basic R‐means clustering algorithm, and expounds several existing improvement methods. An improved R‐means clustering algorithm and a clustering analysis model based on R‐means clustering algorithm are proposed, and the corresponding algorithm flow and implementation are given.

Benchmark and application of unsupervised classification approaches for univariate data

Unsupervised machine learning, and in particular data clustering, is a powerful approach for the analysis of datasets and identification of characteristic features occurring throughout a dataset. It is gaining popularity across scientific disciplines and is particularly useful for applications without a priori knowledge of the data structure. Here, we introduce an approach for unsupervised data classification of any dataset consisting of a series of univariate measurements. It is therefore ideally suited for a wide range of measurement types. We apply it to the field of nanoelectronics and spectroscopy to identify meaningful structures in data sets. We also provide guidelines for the estimation of the optimum number of clusters. In addition, we have performed an extensive benchmark of novel and existing machine learning approaches and observe significant performance differences. Careful selection of the feature space construction method and clustering algorithms for a specific measurement type can therefore greatly improve classification accuracies.

Low-Rank Tensor Regularized Fuzzy Clustering for Multiview Data

Since data are collected from a range of sources via different techniques, multiview clustering has become an emerging technique for unsupervised data classification. However, most existing soft multiview clustering methods only consider the pairwise correlations and ignore high-order correlations among multiple views. To integrate more comprehensive information from different views, this article innovates a fuzzy clustering model using the low-rank tensor to address the multiview data clustering problem. Our method first conducts a standard fuzzy clustering on different views of the data separately. Then, the obtained soft partition results are aggregated as the new data to be handled by a Kullback-Leibler (KL) divergence-based fuzzy model with low-rank tensor constraints. The KL divergence function, which replaces the traditional minimized Euclidean distance, can enhance the robustness of the model. More importantly, we formulate fuzzy partition matrices of different views as a third-order tensor. So, a low-rank tensor is introduced as a norm constraint in the KL divergence-based fuzzy clustering to obtain dexterously high-order correlations of different views. The minimization of the final model is convex and we present an efficient augmented Lagrangian alternating direction method to handle this problem. Specially, the global membership is derived by using tensor factorization. The efficiency and superiority of the proposed approach are demonstrated by the comparison with state-of-the-art multiview clustering algorithms on many multiple-view data sets.

Highly Secure Privacy-Preserving Outsourced k-Means Clustering under Multiple Keys in Cloud Computing

Data clustering is the unsupervised classification of data records into groups. As one of the steps in data analysis, it has been widely researched and applied in practical life, such as pattern recognition, image processing, information retrieval, geography, and marketing. In addition, the rapid increase of data volume in recent years poses a huge challenge for resource-constrained data owners to perform computation on their data. This leads to a trend that users authorize the cloud to perform computation on stored data, such as keyword search, equality test, and outsourced data clustering. In outsourced data clustering, the cloud classifies users’ data into groups according to their similarities. Considering the sensitive information in outsourced data and multiple data owners in practical application, it is necessary to develop a privacy-preserving outsourced clustering scheme under multiple keys. Recently, Rong et al. proposed a privacy-preserving outsourced k-means clustering scheme under multiple keys. However, in their scheme, the assistant server (AS) is able to extract the ratio of two underlying data records, and key management server (KMS) can decrypt the ciphertexts of owners’ data records, which break the privacy security. AS can even reduce all data records if it knows one of the data records. To solve the aforementioned problem, we propose a highly secure privacy-preserving outsourced k-means clustering scheme under multiple keys in cloud computing. In this paper, noncolluded cloud computing service (CCS) and KMS jointly perform clustering over the encrypted data records without exposing data privacy. Specifically, we use BCP encryption which has additive homomorphic property and AES encryption to double encrypt data records, where the former cryptosystem prevents CCS from obtaining any useful information from received ciphertexts and the latter one protects data records from being decrypted by KMS. We first define five protocols to realize different functions and then present our scheme based on these protocols. Finally, we give the security and performance analyses which show that our scheme is comparable with the existing schemes on functionality and security.

Broad Graph-Based Non-Negative Robust Continuous Clustering

Recently, the robust continuous clustering (RCC) has been proposed for unsupervised data classification. The RCC algorithm integrates representation learning and clustering by seeking the balance of the distance of data between intra-cluster and inter-cluster. But the inter-cluster distance in RCC highly depend on the pairwise graph of neighbors, which are constructed by the less-expressive original data. This hampers the performance of RCC on complex epically high-dimensional data. Encouraged by the hybrid feature learning and universal approximation capabilities of the broad learning system (BLS), we first propose a broad graph-based robust continuous clustering algorithm to upgrade RCC. The proposed algorithm measures the distance of the pairwise data with the feature learned from the BLS when constructing the graph. Then to further enhance the clustering performance of the RCC on high-dimensional data, we embed the non-negative matrix factorization (NMF) into the broad graph-based RCC algorithm. By resolving the original data into the basis matrix and coefficient matrix and performing the broad graph-based RCC on the coefficient matrix, the proposed approach takes full advantages of the abundant representation from the BLS and the sparsity and interpretability of NMF coefficients. We verified the proposed algorithms on the synthetic dataset, the UCI dataset and some real-world high dimension datasets. All the empirical results show the proposed algorithms outperform the baselines and improve the clustering performance of RCC effectively.

Delineation of ocean surface provinces over a complex marine domain (off SW Iberia): An objective abiotic-based approach

The partition of surface ocean domains is a tool to assess its patterns and driving forces, ideally incorporating its dynamic nature. The area off Southwest Iberian Peninsula (SWIP), located between temperate and subtropical waters, includes oceanic and coastal domains affected by atmospheric and ocean circulation patterns, as well as coastal processes. The objectives of this study were to delineate a dynamic abiotic-based partition of the heterogeneous surface SWIP area into environmental provinces (EPs), evaluating their spatio-temporal distribution and abiotic properties, as well as the biological relevance of the partition. An unsupervised classification was based on nearly 10 years (2002–2011) of satellite- and model-derived data representative of physical, chemical and optical surface environments. Twelve EPs (two coastal, two slope and eight oceanic) presented areal coverage varying along the annual cycle with patterns significantly related to phytoplankton abundance and productivity. Highly significant intra-annual variability in province-specific area coverage and distinctions in the abiotic properties distinguished EPs predominant during cold (autumn–winter) and warm (spring–summer) periods. Areal coverage and abiotic properties of EPs predominant during the cold period presented higher variability in comparison to warm period EPs. During the warm period, the signature of coastal upwelling was evident, particularly over the western slope and coastal EPs. Riverine discharge was also a relevant local control of abiotic properties over coastal EPs. Overall, the spatio-temporal coverage patterns of EPs and abiotic and biotic properties showed a good agreement with previous studies of the SWIP area, particularly along the oceanic and slope sectors. The combination of a broad dataset, derived from products available through operational oceanography programs, and objective unsupervised data classification methods, represented a suitable strategy for characterizing SWIP ocean surface environment. Direct applications of this partition include its use as support for designing sampling strategies, ocean modelling, interpreting environmental and biological patterns, and ecosystem-based management.

Unsupervised data classification using improved biogeography based optimization

Unsupervised data classification (data clustering) is one of the mostly used data analysis methods which groups the unlabeled data into identical clusters (groups). Classical clustering methods do not perform effectively while clustering high dimensional datasets viz micro array datasets. Therefore, a novel clustering method based on Biogeography based optimization is proposed to extend the capabilities of traditional clustering methods. Performance of proposed method has been tested on the four micro-array datasets. Experimental results validate the effectiveness of proposed method.

IJARCCE - Computer and Communication Engineering

Data clustering which is also called as Cluster Analysis is the unsupervised classification of data into various clusters. Clustering is a method of unsupervised learning which is generally implemented by various machine learning techniques. In this paper a specific comparison of three kinds of clustering is introduced and finally the cost function and loss function are calculated. For evaluating any clustering methods, calculation of the error percentage of the concerned method play an important factor. In this paper comparison is carried out on k-means clustering algorithms, hierarchical algorithms and density based algorithms. The main criteria where focus is given while comparing the clustering algorithms are: Scalability, Classes for dealing with noise and extra deposition, different dimensions of high levels etc.

Clustering Techniques in Bioinformatics

Dealing with data means to group information into a set of categories either in order to learn new artifacts or understand new domains. For this purpose researchers have always looked for the hidden patterns in data that can be defined and compared with other known notions based on the similarity or dissimilarity of their attributes according to well-defined rules. Data mining, having the tools of data classification and data clustering, is one of the most powerful techniques to deal with data in such a manner that it can help researchers identify the required information. As a step forward to address this challenge, experts have utilized clustering techniques as a mean of exploring hidden structure and patterns in underlying data. Improved stability, robustness and accuracy of unsupervised data classification in many fields including pattern recognition, machine learning, information retrieval, image analysis and bioinformatics, clustering has proven itself as a reliable tool. To identify the clusters in datasets algorithm are utilized to partition data set into several groups based on the similarity within a group. There is no specific clustering algorithm, but various algorithms are utilized based on domain of data that constitutes a cluster and the level of efficiency required. Clustering techniques are categorized based upon different approaches. This paper is a survey of few clustering techniques out of many in data mining. For the purpose five of the most common clustering techniques out of many have been discussed. The clustering techniques which have been surveyed are: K-medoids, K- means, Fuzzy C-means, Density-Based Spatial Clustering of Applications with Noise (DBSCAN) and Self- Organizing Map (SOM) clustering.

Multi-class Transductive Learning Based on ℓ 1 Relaxations of Cheeger Cut and Mumford-Shah-Potts Model

Recent advances in l 1 optimization for imaging problems provide promising tools to solve the fundamental high-dimensional data classification in machine learning. In this paper, we extend the main result of Szlam and Bresson (Proceedings of the 27th International Conference on Machine Learning, pp. 1039---1046, 2010), which introduced an exact l 1 relaxation of the Cheeger ratio cut problem for unsupervised data classification. The proposed extension deals with the multi-class transductive learning problem, which consists in learning several classes with a set of labels for each class. Learning several classes (i.e. more than two classes) simultaneously is generally a challenging problem, but the proposed method builds on strong results introduced in imaging to overcome the multi-class issue. Besides, the proposed multi-class transductive learning algorithms also benefit from recent fast l 1 solvers, specifically designed for the total variation norm, which plays a central role in our approach. Finally, experiments demonstrate that the proposed l 1 relaxation algorithms are more accurate and robust than standard l 2 relaxation methods s.a. spectral clustering, particularly when considering a very small number of labels for each class to be classified. For instance, the mean error of classification for the benchmark MNIST dataset of 60,000 data in $\mathbb{R}^{784}$ using the proposed l 1 relaxation of the multi-class Cheeger cut is 2.4 % when only one label is considered for each class, while the error of classification for the l 2 relaxation method of spectral clustering is 24.7 %.

Unsupervised Data Classification for Convex and Non Convex Classes

We present in this work, a new unsupervised data classification technique based on a three steps system: Split, Clean and M erge. In this system, the classes are represented by a set of subclasses that we call prototypes. The prototypes are created in an incremental way from the initial data set. No prior knowledge on the classes is required. The data are presented to the system one by one in an arbitrary way. The system built on a neural network strategy ends up by acquiring knowledge on the data and gathers the data into a set of real classes which may have a non convex structure. The method proposed is compared to the fuzzy C-means ‘FCM’ and fuzzy min max clustering ‘FMMC’ methods through a number of simulations. The results obtained by the proposed method are very good.

Nonlinear dimension reduction and clustering by Minimum Curvilinearity unfold neuropathic pain and tissue embryological classes

Motivation: Nonlinear small datasets, which are characterized by low numbers of samples and very high numbers of measures, occur frequently in computational biology, and pose problems in their investigation. Unsupervised hybrid-two-phase (H2P) procedures—specifically dimension reduction (DR), coupled with clustering—provide valuable assistance, not only for unsupervised data classification, but also for visualization of the patterns hidden in high-dimensional feature space.Methods: ‘Minimum Curvilinearity’ (MC) is a principle that—for small datasets—suggests the approximation of curvilinear sample distances in the feature space by pair-wise distances over their minimum spanning tree (MST), and thus avoids the introduction of any tuning parameter. MC is used to design two novel forms of nonlinear machine learning (NML): Minimum Curvilinear embedding (MCE) for DR, and Minimum Curvilinear affinity propagation (MCAP) for clustering.Results: Compared with several other unsupervised and supervised algorithms, MCE and MCAP, whether individually or combined in H2P, overcome the limits of classical approaches. High performance was attained in the visualization and classification of: (i) pain patients (proteomic measurements) in peripheral neuropathy; (ii) human organ tissues (genomic transcription factor measurements) on the basis of their embryological origin.Conclusion: MC provides a valuable framework to estimate nonlinear distances in small datasets. Its extension to large datasets is prefigured for novel NMLs. Classification of neuropathic pain by proteomic profiles offers new insights for future molecular and systems biology characterization of pain. Improvements in tissue embryological classification refine results obtained in an earlier study, and suggest a possible reinterpretation of skin attribution as mesodermal.Availability: https://sites.google.com/site/carlovittoriocannistraci/homeContact: kalokagathos.agon@gmail.com; massimo.alessio@hsr.itSupplementary information: Supplementary data are available at Bioinformatics online.

Pairwise clustering using a Monte Carlo Markov Chain

In this work an application of MCMC is proposed for unsupervised data classification, in conjunction with a novel pairwise objective function, which is shown to work well in situations where clusters to be identified have a strong overlap, and the centroid oriented methods (such as K-means) fail by construction. In particular, an exceptionally simple but difficult situation is addressed when cluster centroids coincide, and one can differentiate between the clusters only on the basis of their variance. Performance of the proposed approach is tested on synthetic and real datasets.

Density based fuzzy c-means clustering of non-convex patterns

We propose a new technique to perform unsupervised data classification (clustering) based on density induced metric and non-smooth optimization. Our goal is to automatically recognize multidimensional clusters of non-convex shape. We present a modification of the fuzzy c-means algorithm, which uses the data induced metric, defined with the help of Delaunay triangulation. We detail computation of the distances in such a metric using graph algorithms. To find optimal positions of cluster prototypes we employ the discrete gradient method of non-smooth optimization. The new clustering method is capable to identify non-convex overlapped d-dimensional clusters.

Mixture modelling of medical magnetic resonance data

AbstractIn clinical decision making, (semi‐)automatic unsupervised classification of data for diagnostic purposes is becoming more and more important. This paper describes the application of mixture modelling, a clustering where multivariate Gaussians are used to describe clusters in the data, to in vivo nuclear magnetic resonance data of patients with brain tumours. Images as well as localized spectra are analysed. The method is able to automatically generate meaningful classifications. Moreover, the results of clustering both the image and spectral data are in close agreement. Copyright © 2002 John Wiley & Sons, Ltd.