A Unified View of Causal and Non-causal Feature Selection

Author:

Yu Kui1,Liu Lin2,Li Jiuyong2

Affiliation:

1. Hefei University of Technology, Hefei, China

2. University of South Australia, Mawson Lakes, Adelaide, SA, Australia

Abstract

In this article, we aim to develop a unified view of causal and non-causal feature selection methods. The unified view will fill in the gap in the research of the relation between the two types of methods. Based on the Bayesian network framework and information theory, we first show that causal and non-causal feature selection methods share the same objective. That is to find the Markov blanket of a class attribute, the theoretically optimal feature set for classification. We then examine the assumptions made by causal and non-causal feature selection methods when searching for the optimal feature set, and unify the assumptions by mapping them to the restrictions on the structure of the Bayesian network model of the studied problem. We further analyze in detail how the structural assumptions lead to the different levels of approximations employed by the methods in their search, which then result in the approximations in the feature sets found by the methods with respect to the optimal feature set. With the unified view, we can interpret the output of non-causal methods from a causal perspective and derive the error bounds of both types of methods. Finally, we present practical understanding of the relation between causal and non-causal methods using extensive experiments with synthetic data and various types of real-world data.

Funder

Anhui Province Key Research and Development Plan

Australian Research Council Discovery (ARC) Projects

National Natural Science Foundation of China

National Key Research and Development Program of China

Open Project Foundation of Intelligent Information Processing Key Laboratory of Shanxi Province

Publisher

Association for Computing Machinery (ACM)

Subject

General Computer Science

Reference64 articles.

1. Kevin Bache and Moshe Lichman. 2013. UCI machine learning repository. Retrieved from http://archive.ics.uci.edu/ml. Kevin Bache and Moshe Lichman. 2013. UCI machine learning repository. Retrieved from http://archive.ics.uci.edu/ml.

2. On the Feature Selection Criterion Based on an Approximation of Multidimensional Mutual Information

Cited by 51 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Online Learning for Data Streams With Incomplete Features and Labels;IEEE Transactions on Knowledge and Data Engineering;2024-09

2. A novel ensemble causal feature selection approach with mutual information and group fusion strategy for multi-label data;International Journal of Intelligent Computing and Cybernetics;2024-07-22

3. Causal Forest Machine Learning Analysis of Parkinson’s Disease in Resting-State Functional Magnetic Resonance Imaging;Tomography;2024-06-06

4. Discovering causally invariant features for out-of-distribution generalization;Pattern Recognition;2024-06

5. Stable Soft Sensor Modeling for Industrial Systems;2024 IEEE 7th International Conference on Industrial Cyber-Physical Systems (ICPS);2024-05-12

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3