Few-Shot Learning in AI for Science
Organizers
Speaker
Time
Friday, March 21, 2025 3:00 PM - 4:30 PM
Venue
A3-2a-302
Online
Zoom 637 734 0280
(BIMSA)
Abstract
In the current field of AI-assisted scientific research (AI for Science), particularly in drug discovery and biomedicine, we often face the challenge of scarce labeled data. Few-shot learning has become a key technology to address this challenge, as it can effectively leverage limited data for learning and prediction. In this report, I will introduce a series of machine learning algorithms developed specifically to improve data efficiency and prediction accuracy in AI for Science under data scarcity. I will discuss the application of few-shot learning techniques in molecular property prediction, reviewing existing technologies and presenting our proposed Property-Aware Relationship Network (PAR) (NeurIPS 2021, TPAMI 2024) and parameter-efficient Graph Neural Network Adapter (PACIA) (IJCAI 2024). PAR optimizes the relationship representations between molecules by introducing a property-aware molecular encoder and a dependency-query-based relational graph learning module, thereby improving prediction accuracy for various chemical properties. Meanwhile, PACIA enhances few-shot molecular property prediction performance by generating a small number of adaptive parameters to modulate the information propagation process in graph neural networks. In addition, I will introduce the KnowDDI technique (Communications Medicine 2024), which enhances drug representations by leveraging large biomedical knowledge graphs and explains predicted drug-drug interactions (DDIs) by learning knowledge subgraphs of drug pairs, effectively addressing the issue of scarce known data. KnowDDI not only improves prediction performance but also enhances the interpretability of the model, making the prediction process more transparent and trustworthy. Finally, I will share the vision of applying few-shot learning techniques in broader scientific research.
Speaker Intro
Dr. Yaqing Wang is currently an Associate Professor at the Beijing Institute of Mathematical Sciences and Applications. She received her Ph.D. in Computer Science and Engineering from the Hong Kong University of Science and Technology in 2019, under the supervision of Professor Lionel M. Ni and Prof. James T. Kwok. Her research focuses on machine learning. From 2019 to 2024, she worked as a Staff Researcher at Baidu Research, specializing in cold-start recommendation with sparse labeled samples, query understanding, large model and agent optimization, and AI4Science. Her research spans machine learning and artificial intelligence, with a focus on parsimony learning, including few-shot learning, sparse learning, and low-rank learning, aimed at solving real-world problems in biomedicine, recommendation systems, and natural language processing efficiently and cost-effectively. Dr. Wang has published many papers in top-tier international conferences and journals, including NeurIPS, ICML, ICLR, KDD, TheWebConf, SIGIR, EMNLP, TPAMI, JMLR, and TIP, with more than 4000 citations. Her survey on few-shot learning is the most cited paper in ACM Computing Surveys in the past five years and has been recognized as an ESI Hot Paper (top 0.1%). As a key member, she has led major projects such as the National Key R&D Program (Science and Technology Innovation 2030) and projects funded by the National Natural Science Foundation of China. Dr. Wang also serves as a Senior Program Committee member for IJCAI and AAAI and reviews for top conferences and journals, including ICML, NeurIPS, ICLR, and TPAMI. Dr. Wang has been selected for the World's Top 2% Scientists List (single year) by Stanford University and Elsevier company on 2024.