An Empirical Study on Feature Discretization
Machine Learning
2020-04-28 v1 Information Retrieval
Machine Learning
Abstract
When dealing with continuous numeric features, we usually adopt feature discretization. In this work, to find the best way to conduct feature discretization, we present some theoretical analysis, in which we focus on analyzing correctness and robustness of feature discretization. Then, we propose a novel discretization method called Local Linear Encoding (LLE). Experiments on two numeric datasets show that, LLE can outperform conventional discretization method with much fewer model parameters.
Cite
@article{arxiv.2004.12602,
title = {An Empirical Study on Feature Discretization},
author = {Qiang Liu and Zhaocheng Liu and Haoli Zhang},
journal= {arXiv preprint arXiv:2004.12602},
year = {2020}
}