Hyperparameter selection of one-class support vector machine by self-adaptive data shifting

Abstract

With flexible data description ability, one-class Support Vector Machine (OCSVM) is one of the most popular and widely-used methods for one-class classification (OCC). Nevertheless, the performance of OCSVM strongly relies on its hyperparameter selection, which is still a challenging open problem due to the absence of outlier data. This paper proposes a fully automatic OCSVM hyperparameter selection method, which requires no tuning of additional hyperparameter, based on a novel self-adaptive "data shifting" mechanism: Firstly, by efficient edge pattern detection (EPD) and "negatively" shifting edge patterns along the negative direction of estimated data density gradient, a constrained number of high-quality pseudo outliers are self-adaptively generated at more desirable locations, which readily avoids two major difficulties in previous outlier generation methods. Secondly, to avoid time-consuming cross-validation and enhance robustness to noise in the given training data, a pseudo target set is generated for model validation by "positively" shifting each given target datum along the positive direction of data density gradient. Experiments on synthetic and benchmark datasets demonstrate the effectiveness of the proposed method.

MORPHIOUS: A Machine Learning Workflow to Naively Detect the Activation of Microglia and Astrocytes

“…Considering the fact that we do not have a ground truth for activated glia, we cannot rule out that we are over-or under-classifying the activation of glial cells. Tuning hyperparameters for oneclass support vector machines is a critical, and often difficult task, for which a consensus on optimal methodology has yet to be reached 35 . A common technique is to maximize accuracy, while minimizing the number of false-positives, based labeled data (i.e., the class of the data is known) 33,34 .…”

“…outlier data) can often be difficult to acquire. More advanced methods deploy a variety of strategies which focus on identifying patterns in the one-class itself in order to maximize the capacity to distinguish normal cases from outliers 35 .…”

Pan-cancer detection of driver genes at the single-patient resolution

“…Second, we selected the optimal sets of parameters in each kernel. Hyperparameter choice is known to have substantial impacts on classification but it is an open problem for one-class SVMs 22 . Since the parameters for each kernel needed to be selected separately (Supplementary Note), we could not use AUC for assessment.…”

Discriminatory Target Learning: Mining Significant Dependence Relationships from Labeled and Unlabeled Data

“…With the rapid development of computer technologies, business and government organizations create large amounts of data, which need to be processed and analyzed. Over the past decade, to satisfy the urgent need of mining knowledge hidden in the data, numerous machine learning models [1,2] (e.g., decision tree [3], Bayesian network [4,5], support vector machine [6] and Neural network [7]) have been proposed.…”