Adaptive truncated residual regression for finegrained regression problems
[edit]
Proceedings of The Eleventh Asian Conference on Machine Learning, PMLR 101:868882, 2019.
Abstract
Recently, anchorbased regression methods have been applied to challenging regression problems, e.g., object detection and distance estimation, and greatly improved those performances. The key idea of anchorbased regression is to solve the regression of the residuals between selected anchors and original target variable, where the variance is expected to be smaller. However, similar to an ordinary regression method, the anchorbased regression could face difficulty on a finegrained regression and illposed problems where the residual variables tend to be too small and complicated to accurately predict. To overcome these problems on the anchorbased regression, we propose to introduce an adaptive residual encoding in which the too small residual is magnified, and the toolarge residual is truncated using adaptively tuned sigmoidal function. Our proposed method, called ATRNets (Adaptive Truncated ResidualNetworks) with an endtoend architecture could control the range of the target residual to be fitted based on the regression performance, Through experiments with toydata and the system identification for earthquake asperity models, we show the effectiveness of our proposed method.
Related Material


