Victor Fragoso, Walter J. Scheirer, João P. Hespanha et al. · arXiv (Cornell University) · 2016
This work introduces the one-class slab SVM (OCSSVM), a one-class classifier that aims at improving the performance of the one-class SVM. The proposed strategy reduces the false positive rate and increases the accuracy of detecting instances from novel classes. To this end, it uses two parallel hyperplanes to learn the normal region of the decision scores of the target class. OCSSVM extends one-class SVM since it can scale and learn non-linear decision functions via kernel methods. The experiments on two publicly available datasets show that OCSSVM can consistently outperform the one-class SVM and perform comparable to or better than other state-of-the-art one-class classifiers.
Cited by 1Open access
Xiaoxi Zhao, Yingjie Tian, Chichao Zheng · Neural Networks · 2025
Cited by 4
Young-Sik Choi · Pattern Recognition Letters · 2009
Cited by 106
Fa Zhu, Jian Yang, Cong Gao et al. · Neurocomputing · 2016
Cited by 84
Naimul Khan, Riadh Ksantini, Imran Shafiq Ahmad et al. · Figshare · 2023
<p>In one-class classification, the low variance directions in the training data carry crucial information to build a good model of the target class. Boundary-based methods like One-Class Support Vector Machine (OSVM) preferentially separates the data from outliers along the large variance directions. On the other hand, retaining only the low variance directions can result in sacrificing some initial properties of the original data and is not desirable, specially in case of limited training samples. This paper introduces a Covariance-guided One-Class Support Vector Machine (COSVM) classification method which emphasizes the low variance projectional directions of the training data without compromising any important characteristics. COSVM improves upon the OSVM method by controlling the direction of the separating hyperplane through incorporation of the estimated covariance matrix from the training data. Our proposed method is a convex optimization problem resulting in one global optimum solution which can be solved efficiently with the help of existing numerical methods. The method also keeps the principal structure of the OSVM method intact, and can be implemented easily with the existing OSVM libraries. Comparative experimental results with contemporary one-class classifiers on numerous artificial and benchmark datasets demonstrate that our method results in significantly better classification performance.</p>
Cited by 1Open access
Xiaoming Wang, Yong Tian, Xiaohuan Yang et al. · Lecture notes in computer science · 2015
Cited by 1
Naimul Khan, Riadh Ksantini, Imran Shafiq Ahmad et al. · 2023
<p>In one-class classification, the low variance directions in the training data carry crucial information to build a good model of the target class. Boundary-based methods like One-Class Support Vector Machine (OSVM) preferentially separates the data from outliers along the large variance directions. On the other hand, retaining only the low variance directions can result in sacrificing some initial properties of the original data and is not desirable, specially in case of limited training samples. This paper introduces a Covariance-guided One-Class Support Vector Machine (COSVM) classification method which emphasizes the low variance projectional directions of the training data without compromising any important characteristics. COSVM improves upon the OSVM method by controlling the direction of the separating hyperplane through incorporation of the estimated covariance matrix from the training data. Our proposed method is a convex optimization problem resulting in one global optimum solution which can be solved efficiently with the help of existing numerical methods. The method also keeps the principal structure of the OSVM method intact, and can be implemented easily with the existing OSVM libraries. Comparative experimental results with contemporary one-class classifiers on numerous artificial and benchmark datasets demonstrate that our method results in significantly better classification performance.</p>
Cited by 0
Wang, Siting · University of Alberta Library · 2023
Generative Adversarial Networks (GAN) are a field of popular generating models, and there are many variants these years. Lim and Ye 2017 proposed the Geometric GAN to connect the network with the geometric interpretation. They update the discriminator based on the algorithm of Support Vector Machines (SVM), Inspired by their work, we proposed a new algorithm using the robust One Class Support Vector Machines. We also proposed that the discriminator should separate the dataset into three groups: the correctly classified real data, the correctly classified generated data, and the incorrectly classified data. By eliminating the space of incorrectly classified data, we can have our discriminator capture more patterns. We tested our model and the Geometric GAN on the MNIST dataset, and our model has better performance in the same setting.
Cited by 0Open access
Vincent Leclère, Édouard Grave, Laurent El Ghaoui · HAL (Le Centre pour la Communication Scientifique Directe) · 2016
Classification is one of the main problem addressed by machine learning algorithms. Among them the Support Vector Machine (SVM) has attracted a lot of interest and shown success in the past decades. SVM are originally tailored for binary classification. If we have only a few example of negative dataset we can turn to one-class SVM. In this paper we propose a probabilistic interpretation of the one-class SVM approach and an extension especially adapted in the case of highly imbalanced dataset. Indeed, we consider a binary classification problem where we represent the negative dataset by its two first moments, while still modeling the positive class by individual examples. The optimization problem is shown to have an equivalent formulation to a one-class SVM applied to the positive dataset after some preprocess-ing. The usual one-class SVM corresponding to the case where the negative class has mean 0 and identity variance. We show empirically, on a protein classification task and a text classification task, that our approach achieves similar statistical performance than the two mainstream approaches to imbalanced classification problems , while being more computationally efficient .
Cited by 1Open access
Hong-Jie Xing, Lifei Li · Pattern Recognition Letters · 2020
Cited by 21
Shahrani Shahbudin, Masturah Zamri, Murizah Kassim et al. · 2017
Weed classification a necessity in identifying species of weeds to control management practice in agricultural systems, which are essential for maintaining crop productivity and quality. Many classification techniques were used to identify weeds based on images, and most of the techniques using a binary Support Vector Machine (SVM) for measuring the percentage of accuracy. No visualization of decision boundary is illustrated to prove the best performances. To analyzing weed pattern images using One Class Support Vector Machine (SVM), feature vectors of weed images extracted using Gabor Wavelet and Fast Fourier Transform (FFT) were applied. The decision boundaries of the combination extracted feature vectors are visualized and optimal feature vectors are identified. The proposed method also improve the accuracy rate in weed classification task.
Cited by 13
Yanshan Xiao, G.Q. Pan, Bo Liu et al. · Neurocomputing · 2023
Cited by 14
Yuxing Qian, Aimin Feng · 2015
SOCSVM, as the derivative algorithm of OCSVM, is powerful in One-Class Classification. But in big data era, the computing and the storage requirement increase rapidly with the number of training vectors, putting many practical problems out of reach. For applying SOCSVM to large scale data mining, parallel SOCSVM is proposed by means of the idea of MapReduce and the iterative Hadoop model. The experiment results show SOCSVM on Hadoop is efficiency on practical problems.
Cited by 0Open access
Young-Sik Choi, KiJoo Kim · Lecture notes in computer science · 2004
Cited by 2
Peifeng Liang, Weite Li, Yudong Wang et al. · 2018
This paper proposes a novel method for one-class classification by using support vector machine (SVM) based on a divide-and-conquer strategy. An s% winner-take-all autoencoder is applied to realize a sophisticated partitioning which divides the dataset into many clusters. For each cluster, data points are separated from the origin in the feature space like a traditional one-class SVM (OCSVM). By designing a gated linear network, and generating the gate signal from the autoencoder, the proposed OCSVM is implemented in an exact same way as a standard OCSVM with a quasi-linear kernel composed by using a base kernel with the gate signals. Comparing to a traditional OCSVM, the proposed quasi-linear OCSVM is expected to capture a more compact region in the input space. The compact region will decrease the probability of outlier objects falling inside the domain of classifier, which give a better performance. The proposed quasi-linear OCSVM method is applied to different real-world datasets, and simulation results confirm the effectiveness of the proposed method.
Cited by 2
Yu Zhao, Bing Li, Xiu Li et al. · Lecture notes in computer science · 2005
Cited by 101
Jian Luo, Ye Tian, Xin Yan · Soft Computing · 2016
Cited by 13
Ahmed M. Mahfouz, Abdullah Abuhussein, Deepak Venugopal et al. · Algorithms for intelligent systems · 2020
Cited by 15
Harrison O’Neill, Yousaf Ali Khalid, G. G. Spink et al. · Digital Chemical Engineering · 2023
In industrial processes, control valve stiction is known to be one of the primary causes for poor control loop performance. Stiction introduces oscillatory behaviour in the process, leading to increased energy consumption, variations in product quality, shortened equipment lifespan and a reduction in overall plant profitability. Several detection algorithms using routine operating data have been developed over the last few decades. However, with the exception of a handful of recent publications, few attempts to apply classical supervised learning techniques have been published thus far. In this work, principal component analysis, linear discriminant analysis and a one-class support vector machine are trained to detect stiction using time series features as input. These features are extracted from the data using the tsfresh package for Python. The training data consists of simulated stiction examples generated using the XCH stiction model as well as other sources of oscillation. The classifier is subsequently benchmarked against closed-loop stiction data collected in an industrial setting, with performance exceeding that of existing methods.
Cited by 6Open access
Ok, Efkan · Istanbul Technical University Academic Open Archive (Istanbul Technical University) · 2020
Support vector machines (SVMs) are a set of supervised machine learning methods that have recently been applied for structural damage detection. In this thesis, the one-class SVM that requires data only from the undamaged state of the structure for which the damaged data is not available a priori is utilized to train a model for damage detection. The proposed one-class kernel classifier utilizes the autoregressive model with exogeneous input (ARX) parameters and modal frequencies of the intact structure as the training feature to detect structural damage. The trained classifier can then be used with the data obtained from the same structure at different states to classify the structure as damaged or undamaged. The proposed methodology is evaluated using simulated acceleration data of one node from each floor via a finite element model of three-story moment resisting frame under various damage cases. In this study the features selected as damage sensitive feature are ARX model coefficients of related nodes, residual mean errors between prediction accelerations and accelerations obtained from finite element model of related nodes, standard deviations of residual errors and natural frequencies of system. The acceleration signals obtained from sensor locations are processed using Eigensystem Realization algorithm to extract the modal properties of the structure. Reduction in elastic modulus of material and reduction in rigidity of six different connections of structural elements are employed as damage case. The results show that success of model in damage detection varies depending on structural feature used and damage case. While highest damage detection ratios were obtained from SVM-model created with modal frequencies, the performance of SVM-model created with ARX model parameters is also satisfactory.
Cited by 0Open access