Non-Relevance Feedback Document Retrieval based on One Class SVM and SVDD

The SVDD (support vector data description) is one of the most well-known one-class support vector learning methods, in which one tries the strategy of utilizing balls defined on the feature space in order to distinguish a set of normal data from all other possible abnormal objects. The usual strategy of the SVDD depends on the process of finding the region for the normal-class training data with somewhat neglecting the distribution of the abnormal data, thus it may not work well if applied to the binary classification problems in which two classes have similar number of data. In this paper, we consider the problem of performing binary classification based on the SVDD techniques, and in order to overcome the possible drawback of the usual SVDD strategy focusing on the normal-class data only, we propose a new SVDD algorithm which is based on the use of two different SVDD balls for the positive and negative classes along with the SVDD projection and nearest neighbor rule. To investigate how the proposed method works, we compared the performance of the proposed method with SVC (support vector classifier) and conventional SVDD using several real datasets.

Solving one-class problem with outlier examples by SVM

Monitoring of solid-state fermentation of wheat straw in a pilot scale using FT-NIR spectroscopy and support vector data description

A comparative investigation of data-driven approaches based on one-class classifiers for condition monitoring of marine machinery system

Power electronic circuits (PECs) are prone to various failures, whose classification is of paramount importance. This paper presents a data-driven based fault diagnosis technique, which employs a support vector data description (SVDD) method to perform fault classification of PECs. In the presented method, fault signals (e.g. currents, voltages, etc.) are collected from accessible nodes of circuits, and then signal processing techniques (e.g. Fourier analysis, wavelet transform, etc.) are adopted to extract feature samples, which are subsequently used to perform offline machine learning. Finally, the SVDD classifier is used to implement fault classification task. However, in some cases, the conventional SVDD cannot achieve good classification performance, because this classifier may generate some so-called refusal areas (RAs), and in our design these RAs are resolved with the one-against-one support vector machine (SVM) classifier. The obtained experiment results from simulated and actual circuits demonstrate that the improved SVDD has a classification performance close to the conventional one-against-one SVM, and can be applied to fault classification of PECs in practice.

Support vector data description (SVDD) is a well-known kernel-based one-class classification method that exhibits intrinsic regularization ability and robustness versus low numbers of high-dimensional samples. However, the efficiency of SVDD is limited by the cubic time complexity. To solve this problem, this paper first investigates the effect of selecting a reduced subset as the training set of SVDD, while guaranteeing the classification quality. To this end, a new heuristic sample condensed rule, termed HSC, is proposed to accurately identify those potential support vectors that characterize the classification boundary. HSC can consider both the spatial distribution and local density features of training samples, and focus on selecting samples very close to the decision boundary. When dealing with the local density computation, we introduce the idea of K nearest neighbors (KNN) to examine the density of samples in the neighbors of the object to be classified. Finally, a condensed but informative subset obtained by HSC will be applied to train SVDD breezily. The experimental results show that HSC-based SVDD sensibly improves over conventional SVDD, in terms of the size of the training set while guaranteeing a comparable classification quality. In addition, it is competitive over other improved SVDD classifiers in terms of training and testing time.

Anomaly detection using improved deep SVDD model with data structure preservation

Surface electromyography (sEMG) signals are now commonly used in continuous myoelectric control of prostheses. More recently, researchers have considered EMG-based gesture recognition systems for human computer interaction research. These systems instead focus on recognizing discrete gestures (like a finger snap). The majority of works, however, have focused on improving multi-class performance, with little consideration for false activations from "other" classes. Consequently, they lack the robustness needed for real-world applications which generally require a single motion class such as a mouse click or a wake word. Furthermore, many works have borrowed the windowed classification schemes from continuous control, and thus fail to leverage the temporal structure of the gesture. In this paper, we propose a verification-based approach to creating a robust EMG wake word using one-class classifiers (Support Vector Data Description, One Class-Support Vector Machine, Dynamic Time Warping (DTW) & Hidden Markov Models). The area under the ROC curve (AUC) is used as a feature optimization objective as it provides a better representation of the verification performance. Equal error rate (EER) and AUC are then used as evaluation metrics. The results are computed using both window-based and temporal classifiers on a dataset consisting of five different gestures, with a best EER of 0.04 and AUC of 0.98, recorded using a DTW scheme. These results demonstrate a design framework that may benefit the development of more robust solutions for EMG-based wake words or input commands for a variety of interactive applications.

Support vector data description (SVDD) is an effective algorithm for nonlinear process monitoring. However, the conventional SVDD method can't deal with the incipient faults well, which has the small fault amplitude and is easy to be overlapped by industrial noises. Aiming at this problem, an improved SVDD, called probability related SVDD in local variable field (LVPSVDD), is proposed to detect the incipient faults in nonlinear processes. Firstly, the method divides process variables into several local variable fields by hierarchical clustering, and the SVDD model of each variable field is built. Then the sliding window technology is applied to the SVDD distance statistics, and the probability distribution change in the sliding window is measured by Kullback Leibler divergence (KLD). For each local variable field, the corresponding probability related monitoring statistic is developed to replace the original distance statistic. Finally, the global monitoring statistic is obtained by integrating the monitoring results of all local variable fields with Bayesian inference strategy. The method is illustrated with simulation on the continuous stirred tank reactor (CSTR).

Density weighted support vector data description

Support vector data description (SVDD) aims at constructing an optimal hypersphere regarded as a data description for a dataset while support vector classification (SVC) aims at separating data of two classes without providing a data description. This paper proposes a unified approach to both SVDD and SVC that aims at separating data of two classes and at the same time provides a data description. A trade off parameter is introduced to control the balance between describing the data and maximising the margin. Experimental results are provided to evaluate the proposed approach.

As one classical anomaly detection technology, support vector data description (SVDD) has been successfully applied to nonlinear chemical process monitoring. However, the basic SVDD model cannot achieve a satisfactory fault detection performance in the complicated cases because of its intrinsic shallow learning structure. Motivated by the deep learning theory, one improved SVDD method, called ensemble deep SVDD (EDeSVDD), is proposed in order to monitor the process faults more effectively. In the proposed method, a deep support vector data description (DeSVDD) framework is firstly constructed by introducing the deep feature extraction procedure. Different to the traditional SVDD with only one feature extraction layer, DeSVDD is designed with multi-layer feature extraction structure and optimized by minimizing the data-enclosing hypersphere with the regularization of the deep network weights. Further considering the problem that DeSVDD monitoring performance is easily affected by the model structure and the initial weight parameters, an ensemble DeSVDD (EDeSVDD) is presented by applying the ensemble learning strategy based on Bayesian inference. A series of DeSVDD sub-models are generated at the parameter level and the structure level, respectively. These two levels of sub-models are integrated for a holistic monitoring model. To identify the cause variables for the detected faults, a fault isolation scheme is designed by applying the distance correlation coefficients to measure the nonlinear dependency between the original variables and the holistic monitoring index. The applications to the Tennessee Eastman process demonstrate that the proposed EDeSVDD model outperforms the traditional SVDD model and the DeSVDD model in terms of fault detection performance and can identify the fault cause variables effectively.

It is difficult to fully collect foreign body sample in detecting foreign bodies from tobacco.A detection method based on Support Vector Data Description(SVDD) was proposed.Thus a one-class classifier can be developed by using tobacco samples soly.RGB and HSV of tobacco and several typical foreign bodies were firstly extracted;then the HV component was used as eigenvector.A developed SVDD classifier was applied to distinguish foreign bodies from tobacco by inputting the HV eigenvector.Finally through the Receiver Operating Characteristic(ROC) curve,the SVDD classifier was compared with three other methods in classification effect.The experimental results show that by adopting feature extraction with HV component,the data dimension was reduced and a higher computation efficiency was achieved.The SVDD classifier has a stronger classification ability and higher efficiency,which could distinguish foreign bodies from tobacco better.

For the problem of a large dataset, we need to select a subset to represent the original dataset. Many scholars do pattern selection from the problem of the kNN (k-nearest neighbors). The distribution of a pattern's neighbors is usually uneven. In this paper, we define a fixed neighborhood sphere. When the pattern locates near the boundary of the data distribution, there will be fewer neighbors in the neighborhood sphere and when the pattern locates within the data distribution, there will be more neighbors in the neighborhood sphere. According to gather the statistic of the neighbors in a fixed neighborhood sphere, we can find those patterns locating near the boundary of the data distribution. In SVDD (Support Vector Data Description), those patterns are locating near the boundary of the data distribution have more information. They are those patterns which would be support vectors. We can use FNSPS (fixed neighborhood sphere pattern selection) algorithm to select those patterns, which locate near the boundary of the data distribution. The experimental results show that the performance of the SVDD will not go bad. The time complexity of the naive identifying the neighbors in the fixed neighborhood sphere is O(n2). And the time complexity of the SVDD is O(n3). If we set a lower threshold, the FNSPS algorithm can also be used to remove the noise in the targets.

Illumination change is one of most important and difficult problems which prevent from applying face recognition to real applications. For solving this, we propose a method to compensate for different illumination conditions based on SVDD(Support Vector Data Description). In the proposed method, we first consider the SVDD training for the data belonging to the facial images under various illuminations, and model the data region for each illumination as the ball resulting from the SVDD training. Next, we compensate for illumination changes using feature vector projection onto the decision boundary of the SVDD ball. Finally, we obtain the pre-image under the identical illumination with input image. By repeated for each person, we can recognize a person with facial images under same illumination. We also perform the face recognition in order to verify the efficacy of proposed method.