A data reduction approach for resolving the imbalanced data issue in functional genomics

Abstract

Learning from imbalanced data occurs frequently in many machine learning applications. One positive example to thousands of negative instances is common in scientific applications. Unfortunately, traditional machine learning techniques often treat rare instances as noise. One popular approach for this difficulty is to resample the training data. However, this results in high false positive predictions. Hence, we propose preprocessing training data by partitioning them into clusters. This greatly reduces the imbalance between minority and majority instances in each cluster. For moderate imbalance ratio, our technique gives better prediction accuracy than other resampling method. For extreme imbalance ratio, this technique serves as a good filter that reduces the amount of imbalance so that traditional classification techniques can be deployed. More importantly, we have successfully applied our techniques to splice site prediction and protein subcellular localization problem, with significant improvements over previous predictors.

Learning node labels with multi-category Hopfield networks

“…when positive examples are significantly fewer than those negative, and imbalance-aware strategies are required [16,67,35,34]. Unfortunately, several graph-based prediction problems are characterized by strongly unbalanced labelings [13,53,21].…”

Imbalanced classification using support vector machine ensemble

“…Chawla et al [16] proposed the synthetic minority oversampling technique (SMOTE) algorithm in which the minority class was oversampled by taking each minority class sample and introducing new synthetic examples joining any or all the minority class nearest neighbors. Further, algorithms that combine SMOTE and other learning methods have also been applied to solve the class imbalance problem [17][18][19].…”

A supervised learning approach for imbalanced data sets

“…A problem that we focus on this paper is how to learn a classification task from imbalanced data sets. The class imbalance problem, which is one of the fundamental problems in machine learning, has received much attention recently [1,4,9,14,15]. In many real-world diagnostic applications, e.g., computer security, biomedical, and engineering, uneven distribution of data patterns is very common, where number of training instances of a minority class is much smaller compared to other majority classes; as a result, the classifier tends to favor the majority class [7].…”

“…In general, learning algorithms for class imbalance problems can be divided into two categories: resampling and cost-sensitive based. Resampling meth-978-1-4244-2175-6/08/$25.00 ©2008 IEEE ods such as over-sampling and under-sampling [7,14] modify the prior probability of the majority and minority class in the training set to obtain a more balanced number of instances in each class. The undersampling method extracts a smaller set of majority instances while preserving all the minority instances.…”

“…Their method improves the classification accuracy for minority classes by assigning larger cost to them. However, adding a cost function in the learning process will modify the probability distribution [14].…”