Minimum variance embedded auto-associative kernel extreme learning machine for one-class classification

One-class classification (OCC) needs samples from only a single class to train the classifier. Recently, an auto-associative kernel extreme learning machine was developed for OCC task. This paper introduces novel extension of this classifier by embedding minimum variance information within its architecture and is referred as VAAKELM. The forces network output weights focus in regions low reduces intra-class variance. leads better separation target outliers, resulting improvement generalization performance proposed follows reconstruction-based approach minimizes reconstruction error using base It uses deviation identify outliers. We perform experiments on 15 small-size 10 medium-size one-class benchmark datasets demonstrate efficiency compare results with 13 existing classifiers considering mean $$\hbox {F}_1$$ score comparison metric. experimental show that VAAKELM consistently performs than classifiers, making it viable alternative source code available GitHub homepage: https://github.com/PratikMishra/VAAKELM.