Professor Chongsheng Zhang (Department of Data Science, Henan University, China) is going to lecture on
From Imbalance Learning to Long-tailed Learning
at Classroom H3, Mezzanine, Building of Biology Department, University Campus (https://www.csd.auth.gr/en/the-school/access/) on Tuesday October 8th, 2024 at 11 am. The lecture is open to the academic community of Aristotle University.
ABSTRACT
Long-tailed data is a special type of multi-class imbalanced data with a very large amount of minority/tail classes that have a very significant combined influence. Long-tailed learning aims to build high-performance models on datasets with long-tailed distributions, which can identify all the classes with high accuracy, in particular the minority/tail classes. It is a cutting-edge research direction that has attracted a remarkable amount of research effort in the past few years. In this talk, I will present a comprehensive survey of latest advances in long-tailed visual learning. First a new taxonomy for long-tailed learning will be presented, which consists of eight different dimensions, including data balancing, neural architecture, feature enrichment, logits adjustment, loss function, bells and whistles, network optimization, and post hoc processing techniques. Based on the proposed taxonomy, a systematic review of long-tailed learning methods will be undertaken, discussing their commonalities and alignable differences. The differences between imbalance learning and long-tailed learning approaches will be analyzed. Finally, prospects and future directions
About the Speaker:
Chongsheng Zhang obtained his PhD from INRIA, France. Since then, he has been working at Henan University, China, where he is a Professor and leader of the data science and artificial intelligence team. His research interests include imbalance/long-tailed learning, OCR and digital paleography. He has published 50 papers in peer-reviewed journals and conferences, including AAAI, KDD, etc. He has authored 6 books and filed 20 Chinese patents.
in this field will be discussed.
| Chongsheng Zhang
Professor School of Computer and Information Engineering Henan University, China Email: cszhang@ieee.org |
