Twin support vector quantile regression

TL;DR

The paper introduces TSVQR, a novel twin support vector quantile regression method that efficiently captures data heterogeneity and asymmetry, outperforming previous methods in accuracy and speed across diverse datasets.

Contribution

TSVQR is a new quantile regression approach that constructs smaller quadratic programming problems and uses a dual coordinate descent algorithm for faster training.

Findings

Outperforms previous quantile regression methods in accuracy.

Faster training due to smaller QPPs and dual coordinate descent.

Effective on artificial, benchmark, large-scale, time-series, and imbalanced datasets.

Abstract

We propose a twin support vector quantile regression (TSVQR) to capture the heterogeneous and asymmetric information in modern data. Using a quantile parameter, TSVQR effectively depicts the heterogeneous distribution information with respect to all portions of data points. Correspondingly, TSVQR constructs two smaller sized quadratic programming problems (QPPs) to generate two nonparallel planes to measure the distributional asymmetry between the lower and upper bounds at each quantile level. The QPPs in TSVQR are smaller and easier to solve than those in previous quantile regression methods. Moreover, the dual coordinate descent algorithm for TSVQR also accelerates the training speed. Experimental results on six artiffcial data sets, ffve benchmark data sets, two large scale data sets, two time-series data sets, and two imbalanced data sets indicate that the TSVQR outperforms previous quantile regression methods in terms of the effectiveness of completely capturing the heterogeneous and asymmetric information and the efffciency of the learning process.