A Quadratic Mean based Supervised Learning Model for Managing Data Skewness

Abstract

In this paper, we study the problem of data skewness. A data set is skewed/imbalanced if its dependent variable is asymmetrically distributed. Dealing with skewed data sets has been identified as one of the ten most challenging problems in data mining research. We address the problem of class skewness for supervised learning models which are based on optimizing a regularized empirical risk function. These include both classification and regression models for discrete and continuous dependent variables. Classical empirical risk minimization is akin to minimizing the arithmetic mean of prediction errors, in which approach the induction process is biased towards the majority class for skewed data. To overcome this drawback, we propose a quadratic mean based learning framework (QMLearn) that is robust and insensitive to class skewness. We will note that minimizing the quadratic mean is a convex optimization problem and hence can be efficiently solved for large and high dimensional data. Comprehensive experiments demonstrate that the QMLearn model significantly outperforms existing statistical learners including logistic regression, support vector machines, linear regression, support vector regression and quantile regression etc.

Classification of Passes in Football Matches Using Spatiotemporal Data

“…The function (4) is the risk function derived from the maximum likelihood estimation of θ. The arithmetic (5) and quadratic (6) risk functions are intended to perform better under class imbalance conditions by computing the per-class risk [14]. The arithmetic risk takes the sum of the per-class values, whereas the quadratic risk uses the root of the sum of the squared values.…”

Targeted Attention Attack on Deep Learning Models in Road Sign Recognition

“…[8], and are likely to enable the future self-driving cars [9]. However, recent studies have shown that DNNs tend to deliver a wrong prediction with high confidence if the input data are slightly perturbed [10][11] [12][13] [14]. These perturbations are often imperceptible to human vision system, but their effects on DNNs are catastrophic [15].…”

Learning to Propagate Rare Labels

“…If the input data points were not connected in the form of a graph, the problem could have been formulated as either semi-supervised outlier detection [12] or supervised classification from imbalanced data [6,16,5,19,15,11] depending on whether degree of supervision is low or high. The existing literature, however, appears to be quite slim when it comes to address the problem of semi-supervised graph label propagation for rare labels.…”