Title:"Achieving Self-Optimizing Feature Engineering through Reinforcement Learning"
Speaker:Dr. Kunpeng Liu, Assistant Professor, Department of Computer Science, Portland State University
Date:November 29, 2023 (Wednesday)
Time:10:00 AM - 12:00 PM
Venue:Room 902, Science and Education Building A, Emerald Lake Campus
In recent years, data mining has achieved significant success in many application scenarios. As a fundamental technology in data mining, feature engineering plays an irreplaceable role in the process of understanding and perceiving data. Ideal feature engineering can remove irrelevant features, generate informative features, improve model performance, enhance generalization, and provide better interpretability. However, in many application scenarios, most practitioners are not experts in feature engineering. Therefore, automated feature engineering to lower the threshold of feature engineering becomes an indispensable requirement. This presentation will first introduce the importance and challenges of automated feature engineering, focusing on: 1. Automated feature selection; 2. Automated feature generation, and discuss how to use the framework of reinforcement learning to address these issues. Finally, it will look forward to future intelligent, interpretable, and interactive automated data science systems and propose several development directions.
Dr. Kunpeng Liu is an Assistant Professor in the Department of Computer Science at Portland State University. His research interests lie in data mining and reinforcement learning. His recent research focuses on automated data science systems and their applications in big data problems, including smart cities, machine learning privacy protection, interpretable recommendation systems, and user behavior analysis. His research results have been published in top conferences and journals in data mining and machine learning, including KDD, TKDE, IJCAI, AAAI, WWW, etc. Dr. Liu has served as a senior program committee member for IJCAI and a regular program committee member for top international conferences such as KDD, ICML, ICLR, NeurIPS, AAAI, WWW, CIKM, ICDM.