Report Title: Research on Causal Feature Selection and Structure Learning

Time: 16:10, Saturday, November 1, 2025

Venue: Room 1602, Building A, Science and Education Building, Feicui Lake Campus

Speaker: Associate Professor Zhaolong Ling

Affiliation: Anhui University

Organizer: School of Computer Science and Information Engineering

Report Abstract:

Over the past three decades, feature selection, as a dimensionality reduction technique in big data, has been one of the research hotspots in machine learning. Traditional feature selection algorithms, relying only on correlations between features and class attributes, may lead to prediction and classification models lacking interpretability, operability and robustness. Causal feature selection discovers a substructure of the Bayesian network of the class attribute, namely the Markov blanket, which consists of parents (direct causes), children (direct effects) and spouses (other direct causes of children) of the class attribute, thus explicitly deriving local causal relationships between features and the class attribute. As an emerging feature selection method, causal feature selection has attracted extensive attention in machine learning and causal discovery by identifying potential causal features to build interpretable, operable and robust predictive classification models. Besides classification-oriented feature selection, as a substructure of Bayesian networks, causal feature selection plays a vital role in learning local causal network structures of target variables. Furthermore, if Markov blankets of all variables in a dataset can be identified, they can be used as constraints to reduce the search space, enabling efficient local-to-global full causal network structure learning.

Speaker Profile:

Zhaolong Ling is an Associate Professor, Deputy Director and Master Supervisor at the School of Computer Science and Technology, Anhui University. He received his PhD degree from the School of Computer Science and Information Engineering, Hefei University of Technology in June 2020. He is currently a Director of the Anhui Artificial Intelligence Society, Secretary-General of the Causal and Cognitive Intelligence Committee of Anhui Artificial Intelligence Society, and a member of the Granular Computing and Knowledge Discovery Committee of the Chinese Association for Artificial Intelligence.

His main research interests include data mining and causal reasoning. In recent years, he has published more than 20 papers in high-level international journals (e.g., IEEE TPAMI, IEEE TKDE, ACM TIST, ACM TKDD) and conferences (e.g., IJCAI, AAAI, ACM MM, IJCNN). He serves as a reviewer for high-level international journals including IEEE TKDE, IEEE TNNLS, ACM TIST, IEEE TBD, and a PC member for top conferences including ICML, KDD, IJCAI, AAAI. He has led/participated in 5 projects including the National Key R&D Program, General Program and Youth Program of the National Natural Science Foundation of China.