Green Machine Learning with Knowledge Graphs
- LecturerDr. Yun-Cheng (Joe) Wang (Electrical and Computer Engineering, University of Southern California)
Host: Hsin-Min Wang - Time2023-12-29 (Fri.) 10:30 ~ 12:30
- LocationAuditorium106 at IIS new Building
Abstract
To build an advanced artificial intelligence (AI) system, incorporating transparent and efficient machine learning models is essential. Particularly, green machine learning models target small model sizes, small numbers of computation operations, and interpretable model designs to build sustainable solutions. It is even more attractive to leverage knowledge graphs (KGs), multi-relational graphs that store human knowledge, in green machine learning models to improve the explainability of the predictions. In this talk, I will introduce two green machine learning approaches, GreenKGC and AsyncET, that aim to infer the missing information in the KGs. Specifically, GreenKGC predicts missing links under a lower-dimensional space, while AsyncET improves the embedding quality by incorporating entity types during representation learning.
BIO
Yun-Cheng (Joe) Wang obtained his Ph.D. in Electrical and Computer Engineering at the University of Southern California (USC) in December 2023, supervised by Prof. C.-C. Jay Kuo. Prior to joining USC, he received his B.S. in Electrical Engineering from National Taiwan University in 2018. His research focuses on developing lightweight, efficient, and transparent machine learning (ML) models with specific interests in diverse knowledge graph applications, such as KG completion, KG representation learning, and interpretable and efficient reasoning on KGs.