Separable Data Research Articles

Community Detection Algorithm Based on Nonnegative Matrix Factorization and Improved Density Peak Clustering

Community detection is a critical issue in the field of complex networks. Recently, the nonnegative matrix factorization (NMF) method has successfully uncovered the community structure in the complex networks. However, this method has a significant drawback; most of community detection methods using NMF require the number of communities to be preassigned or determined the number of communities by searching for the best community structure among all candidates. To address this problem, in this paper, we use density peak clustering (DPC) to obtain the number of centers as the pre-defined parameter for nonnegative matrix factorization. However, due to sparse and high dimensional characteristics of complex networks, DPC cannot be used to detect community directly. To overcome this issue, we employ degree and hop of nodes as the density and distance indexes, respectively; we use NMF and Symmetric NMF to deal with linearly separable data and non-linearly separable data, respectively. Experimental results show that the proposed methods exhibit excellent performance on artificial and real-world networks and superior to the state-of-the-art methods which are the most common method for community detection of complex networks.

A separable reversible data hiding scheme in encrypted image for two cloud servers

A separable reversible data hiding scheme in encrypted image for two cloud servers

Human Activity Recognition Based on Parallel Approximation Kernel K-Means Algorithm

Recently, owing to the capability of mobile and wearable devices to sense daily human activity, human activity recognition (HAR) datasets have become a large-scale data resource. Due to the heterogeneity and nonlinearly separable nature of the data recorded by these sensors, the datasets generated require special techniques to accurately predict human activity and mitigate the considerable heterogeneity. Consequently, classic clustering algorithms do not work well with these data. Hence, kernelization, which converts the data into a new feature vector representation, is performed on nonlinearly separable data. This study aims to present a robust method to perform HAR data clustering to mitigate heterogeneity in data with minimal resource consumption. Therefore, we propose a parallel approximated clustering approach to handle the computational cost of big data by addressing noise, heterogeneity, and nonlinearity in data using data reduction, filtering, and approximated clustering methods on parallel computing environments that have not been previously addressed. Our key contribution is to treat HAR as big data implemented by approximation kernel K-means approaches and fill the gap between the HAR clustering cost and parallel computing fields. We implemented our approach on Google cloud on a parallel spark cluster, which helped us to process large-scale HAR data across multiple machines of clusters. The normalized mutual information is used as validation metric to assess the quality of the clustering algorithm. Additionally, the precision, recall, f-score metrics values are obtained somehow to compare the results with a classification technique. The experimental results of our clustering approach prove its effectiveness compared with a classification technique and can efficiently detect physical activity and mitigate the heterogeneity of the datasets.

Community detection algorithm based on nonnegative matrix factorization and pairwise constraints

Community detection is a critical issue in the field of complex networks. Nonnegative matrix factorization (NMF) has been one of the hot research topics in community detection. In the real world, only topology information is inadequate to detect community structure accurately. To utilize the limited supervised information effectively, we propose two algorithms for community detection, which combine NMF or (Symmetric NMF) with pairwise constraints to deal with linearly separable data and nonlinearly separable data, respectively. Currently, most of NMF-based community detection methods need the number of communities in advance or searching among all candidates with different K. To address this issue, we use singular value decomposition to reveal the number of communities automatically. Experimental results demonstrate that the proposed methods can produce higher clustering performance compared with some state-of-the-art methods.

Influence of kernel clustering on an RBFN

Classical radial basis function network (RBFN) is widely used to process the non-linear separable data sets with the introduction of activation functions. However, the setting of parameters for activation functions is random and the distribution of patterns is not taken into account. To process this issue, some scholars introduce the kernel clustering into the RBFN so that the clustering results are related to the parameters about activation functions. On the base of the original kernel clustering, this study further discusses the influence of kernel clustering on an RBFN when the setting of kernel clustering is changing. The changing involves different kernel-clustering ways [bubble sort (BS) and escape nearest outlier (ENO)], multiple kernel-clustering criteria (static and dynamic) etc. Experimental results validate that with the consideration of distribution of patterns and the changes of setting of kernel clustering, the performance of an RBFN is improved and is more feasible for corresponding data sets. Moreover, though BS always costs more time than ENO, it still brings more feasible clustering results. Furthermore, dynamic criterion always cost much more time than static one, but kernel number derived from dynamic criterion is fewer than the one from static.

Arrhythmia Classification with Single Beat ECG Evaluation and Support Vector Machine

Abnormal electrical activity of the human heart indicates cardiac dysfunction. The Electrocardiogram (ECG) is one of the non-invasive diagnostic techniques to detect cardiac abnormalities. Irregularity and non-stationarity in the ECG signal impose difficulties to clinicians for accurate diagnosis of heart diseases only by visual inspection. Automatic recognition of abnormal ECG beats aids in early detection of heart diseases. This paper explores the ECG single beat analysis to identify the cardiac abnormality. In this work, seven classes of arrhythmia are considered as recommended by AAMI(Association for the Advancement of Medical Instrumentation) standard. Beat feature database is generated from 44 recordings of the MIT-BIH arrhythmia database to support the arrhythmia classification. Classification is implemented with Multiclass Support Vector Machine (SVM) for non-linearly separable data effectively. Classification accuracy up to 93% is achieved for the selected input feature sets. This work assesses the suitability of the ECG input features for multi-class classification of arrhythmia.

Separable Reversible Data Hiding in Encrypted Image Based on Two-Dimensional Permutation and Exploiting Modification Direction

In this paper, we propose a separable reversible data hiding method in encrypted image (RDHEI) based on two-dimensional permutation and exploiting modification direction (EMD). The content owner uses two-dimensional permutation to encrypt original image through encryption key, which provides confidentiality for the original image. Then the data hider divides the encrypted image into a series of non-overlapping blocks and constructs histogram of adjacent encrypted pixel errors. Secret bits are embedded into a series of peak points of the histogram through EMD. Direct decryption, data extraction and image recovery can be performed separately by the receiver according to the availability of encryption key and data-hiding key. Different from some state-of-the-art RDHEI methods, visual quality of the directly decrypted image can be further improved by the receiver holding the encryption key. Experimental results demonstrate that the proposed method outperforms some state-of-the-art methods in embedding capacity and visual quality.

Noise Clustering-Based Hypertangent Kernel Classifier for Satellite Imaging Analysis

The classification accuracy and the computational complexity are degraded by the occurrence of nonlinear data and mixed pixels present in satellite images. Therefore, the kernel-based fuzzy classifiers are required for the separation of linear and nonlinear data. This paper presents two classifiers for handling the nonlinear separable data and mixed pixels. The classifiers, noise clustering (NC) and NC with hypertangent kernels (NCH), are used for handling these problems in the satellite images. In this study, a comparative study between NC and NCH has been carried out. The membership values of KFCM are obtained to produce the final result. It is found that the proposed classifiers achieved good accuracy. It is observed that there is an enhancement in the classification accuracy by using NC and NCH. The maximum accuracy achieved for NC and NCH is 75% at δ = 0.7, δ = 0.5, respectively. After comparing both the results, it has identified that NCH gives better results. The classification of Formosat-2 data is done by obtaining optimized values of m and δ to generate the fractional outputs. The classification accuracy is performed by using the error matrix with the incorporation of hard classifier and α-cut.

A review and proposal of (fuzzy) clustering for nonlinearly separable data

In many practical situations data may be characterized by nonlinearly separable clusters. Classical (hard or fuzzy) clustering algorithms produce a partition of objects by computing the Euclidean distance. As such, they are based on the linearity assumption and, therefore, do not identify properly clusters characterized by nonlinear structures. To overcome this limitation, several approaches can be followed: density-, kernel-, graph- or manifold-based clustering. A review of these approaches is offered and some new fuzzy manifold-based clustering algorithms, involving the so-called geodesic distance, are proposed. The effectiveness of such algorithms is shown by synthetic, benchmark and real data.

Jointly Using Low-Rank and Sparsity Priors for Sparse Inverse Synthetic Aperture Radar Imaging.

The inverse synthetic aperture radar (ISAR) imaging technique of a moving target with sparse sampling data has attracted wide attention due to its ability to reduce the data collection burden. However, traditional low-rank or 2D compressive sensing (CS)-based ISAR imaging methods can handle the random sampling or the separable sampling data only. When the specific data collection condition cannot be satisfied, low-rank or 2D CS-based methods cannot provide satisfactory imaging results any more. To remedy this problem, in this paper, we proposed a joint low-rank and sparsity priors' constrained model for ISAR imaging with various sparse data patterns. This model is inspired by the facts that the received radar data have a low-rank property and the ISAR image is sparse on the specific dictionary. Two reconstruction algorithms to solve the double priors' constrained optimization problem are developed under the alternative direction method of multipliers (ADMM) framework with the help of augmented Lagrange multipliers (ALM). Results on simulation data and real data show that the proposed methods are quite effective in recovering missing samples and focused image and perform better than the matrix completion-based method and the sparse representation-based method when dealing with the various kinds of sparse sampling data.

An Efficient Separable Reversible Data Hiding Using Paillier Cryptosystem for Preserving Privacy in Cloud Domain

Reversible data hiding in encrypted image (RDHEI) is advantageous to scenarios where complete recovery of the original cover image and additional data are required. In some of the existing RDHEI schemes, the image pre-processing step involved is an overhead for the resource-constrained devices on the sender’s side. In this paper, an efficient separable reversible data hiding scheme over a homomorphically encrypted image that assures privacy preservation of the contents in the cloud environment is proposed. This proposed scheme comprises three stakeholders: content-owner, data hider, and receiver. Initially, the content-owner encrypts the original image and sends the encrypted image to the data hider. The data hider embeds the encrypted additional data into the encrypted image and then sends the marked encrypted image to the receiver. On the receiver’s side, both additional data and the original image are extracted in a separable manner, i.e., additional data and the original image are extracted independently and completely from the marked encrypted image. The present scheme uses public key cryptography and facilitates the encryption of the original image on the content-owner side, without any pre-processing step involved. In addition, our experiment used distinct images to demonstrate the image-independency and the obtained results show high embedding rate where the peak signal noise ratio (PSNR) is +∞ dB for the directly decrypted image. Finally, a comparison is drawn, which shows that the proposed scheme is an optimized approach for resource-constrained devices as it omits the image pre-processing step.

On the optimal solution to maximum margin projection pursuit

Most existing dimensionality reduction methods have been applied as a separable data preprocessing step before classification algorithms. This reduces the flexibility of classification algorithms. To handle this limitation, recently, a novel method, namely maximum margin projection pursuit (MMPP), was developed by simultaneously taking into account dimensionality reduction and classification in the criterion function. MMPP alternatively updates the projection matrix and normal vector of classifications hyperplane by optimizing the min-max problem. This results in the following two problems: (1) It does not guarantee both the convergence of the proposed iterative algorithm in real applications and minimization of the objective function; (2) It heavily depends on learning rate and does not get the global optimal solution. In this paper, we simultaneously solve both the projection matrix and norm vector of classification hyperplane by non-iterative method which not only optimizes the criterion function but also is faster than traditional MMPP algorithm. Furthermore, we extend our method to solve multiclass classification problems. Experiments on the Yale, ORL, AR and COIL20 databases have been conducted to evaluate our method. The results illustrate that, compared with the iterative algorithm, our no-iteration algorithm achieves higher efficiency, more stable recognition, and smaller objective value.

Separable reversible data hiding in encrypted images based on scalable blocks

This paper proposes a new separable reversible data hiding method for encrypted images. Proposed scheme employs the pixel redundancy of natural images to construct embedding space. First, cover image is divided into multiple blocks with different scales. According to the pixel average value of each block, the lowest two bits of every pixel are vacated as reserved rooms. Subsequently, the whole image is encrypted by using stream cipher and the secret messages are finally embedded into the reserved rooms by the embedding key. Proposed scheme is separable in the sense that the recipient can achieve different function by the following ways: (a) If the recipient has only decryption key, an approximation plaintext image containing the embedded information can be obtained. (b) If the recipient has only embedded key, secret messages can be extracted correctly. (c) If the recipient has both decryption key and embedded key, he can not only extract the secret messages, but recover the original cover image perfectly. Extensive experiments are performed to show that our proposed schemes outperform existing reversible data hiding schemes in terms of visual quality, embedding capacity and security performance, even if a large-scale image database is used.

Stock Market Prediction Using Machine Learning

Stock price prediction is one of the most complex machine learning problems. It depends on a large number of factors which contribute to changes in the supply and demand. In this paper, we propose a stock prediction analysis using machine learning based on support vector machines (SVM), linear regression and reinforcement learning. SVM are favored in applications where text mining is used for market prediction. SVMs can be used for both linearly and non-linearly separable data sets. when the data is linearly separable, SVMs construct a hyperplane on the feature space to distinguish the training tuples in the data such that the margin between the support vectors is maximized. Correlation is used between stock prices of different companies to predict the price of a stock by using technical indicator of highly correlated stocks, not only stock to be predicted.

Cognitive Framework for HIV-1 Protease Cleavage Site Classification Using Evolutionary Algorithm

Acquired immunodeficiency syndrome is the progression from human immunodeficiency virus (HIV) to symptomatic disease. The disease process starts by replicating the viral particles, and the HIV-1 protease is the causative agent behind the phenomenon. To control the replication, protease inhibitors are designed that bind to the active sites. Thus, knowledge about the HIV protease cleavage specificity is critical in the design of effective HIV inhibitors. Several classifiers and encoding techniques have been proposed for the identification of cleavage sites. A recent study shows a linear support vector machine with orthogonal encoding promulgated its sovereignty in the linearly separable peptide data. However, the performance of linear SVM degrades on nonlinear separable data. To cope with the considerable changes between linear and nonlinear separable data, we propose a cognitive framework that utilizes the dexterity of evolutionary algorithm on exploring out the best adaptive settings in the classifier and encoding techniques to improve the predictive capability of the model. Specifically, the proposed cognitive model selects the three decisive factors: physicochemical property for peptide encoding, SVM kernel selection for linear or nonlinear data and parameter tuning of the selected kernel. To validate the performance of the proposed model, the benchmark peptide data are exercised in the experimentation, and the obtained results are compared with the state-of-the-art techniques. Besides, we also analyzed the performance of the proposed model in out-of-sample test, and the extensive experimental results prove the proficiency of the proposed approach for the classification of the HIV-1 protease cleavage sites.

A novel robust kernel for classifying high-dimensional data using Support Vector Machines

This paper presents a new semantic kernel for classification of high-dimensional data in the framework of Support Vector Machines (SVM). SVMs have gained widespread application due to their relatively higher accuracy. The efficacy of SVMs, however, depends upon the separation of the data itself as well as the kernel function. Text data, for instance, is difficult to classify due to synonymy and polysemy in its contents, having multi-topical instances that can result in mislabeling, and being highly sparse in the bag-of-words representation. While the soft margin parameter and kernel tricks are used in SVM to deal with outliers and non-linearly separable data, using data statistics and correlation has not been fully explored in the literature. This paper explore the use co-similarity (i.e., soft co-clustering) to find latent relationships between documents motivated by the success of co-clustering and subspace clustering methods. It has been shown that the use of weighted higher-order paths between instances in the data can be a good measure of similarity values which can then be used for both classification and to correct mislabeled (or outlier) data in the training set. The proposed kernel is generic in nature and suitable for sparse, dyadic data where direct co-occurrences are not necessary common as in the case of textual data, link-analysis in social media networks, co-authorship, etc. It also studies the impact of noise in the training data and provides a technique to re-label such instances. It is also observed that re-labelling of selected training data reduces the adverse effect of outliers or label noise and can greatly improve the classification of the test data. To the best of our knowledge, we are the first to introduce a supervised co-similarity based kernel function and also provide mathematical formulation to show that it is a valid Mercer's kernel. Our experiments show that the proposed framework outperforms current and state-of-the-art methods in terms of classification accuracy and is more resilient to label noise.

Fuzzy DEA-based classifier and its applications in healthcare management.

Nonlinear fuzzy classification models have better classification performance than linear fuzzy classifiers. In many nonlinear fuzzy classification problems, piecewise-linear fuzzy discriminant functions can approximate nonlinear fuzzy discriminant functions. In this paper, we first build fuzzy classifier based on data envelopment analysis (DEA) for incremental separable fuzzy training data, which can be widely applied in the healthcare management with fuzzy attributes, then we apply the proposed fuzzy DEA-based classifier in the diagnosis of Coronary with fuzzy symptoms and the classification of breast cancer dataset with fuzzy disturbance. Numerical experiments show the proposed fuzzy DEA-based classifier is accurate and robust.

An Interpolative AMBTC-based high-payload RDH scheme for encrypted images

A reserving room before encryption (RRBE) framework is proposed to provide separable reversible data hiding for encrypted images in this paper. A combination of Interpolative AMBTC and Huffman coding is used to enhance the hiding capacity. In our scheme, a receiver with only the hiding key can extract the secret data without knowing about the content. If the receiver has only the encryption key, it is not possible to extract the hidden secret data, but it is possible to decrypt an image similar to the original image. Only when the receiver has both the hiding key and the encryption key, it is possible to extract the secret data and completely recover the original content without any error. The experimental results demonstrated the efficiency of our proposed high payload reversible data hiding scheme for encrypted images (HP-RDHEI), especially with respect to the data embedding rate.

A New High Capacity Separable Reversible Data Hiding in Encrypted Images Based on Block Selection and Block-Level Encryption

Reversible data hiding in encrypted image (RDHEI) is a technology that can simultaneously implement privacy protection and reversible information hiding. For RDHEI, the main problem is the conflict between embedding capacity and real reversibility. In this paper, to solve such a conflict, a new block-level image encryption method is proposed to fulfill privacy protection while preserving the spatial correlation of pixels inside pixel block. On this basis, a new reversible data hiding method based on block selection is also proposed. Specifically, pixel blocks are classified into embeddable block (EB) and non-embeddable block (NEB). Then, additional data is embedded into only EBs and the position file locating EB is also embedded using a new self-embedding method. Experimental results demonstrate that the proposed method not only achieves high embedding capacity, but also guarantees lossless data extraction and perfect image recovery.

Performance Comparison of SVM and its Variants for the Early Prognosis of Breast Cancer

Breast cancer has become a leading cause of women death in this era. Breast cancer is very common in various countries including Pakistan. Early identification of the breast cancer or tumor is the only way for the rapid treatment and cure. An imaging approach named as mammography has performed tremendous job in the field of medical to detect the cancer tumors on early basis with less false alarm rate. Breast cancer has two types of tumors a) Benign and b) Malignant. Malignant is acknowledged as cancer tumor as it spread and grow rapidly inside the tissues. Detection of Malignant tumor is very complex in dense breast as it is covered and linked with the milk glands, ducts and other related tissues. Therefore, machine learning and artificial intelligence approaches were needed as mammographic images required edge detection, image enhancement and image processing. Various Artificial Intelligence based algorithms have been applied to the clinical breast cancer data set for the early detection of breast tumor. In this research work the clinical data has been collected from the UCI machine learning repository for the classification of breast cancer tumor a) Benign and b) Malignant. Support Vector machine with its variants Kernel, Gaussian Kernel and Sigmoid Kernel have been applied to the linearly separable breast cancer data set for comparative analysis. Results proved that all the variants of SVM performed better for the breast cancer classification.&nbsp