3 papers
stat.ML2026
Characterizing and Correcting Effective Target Shift in Online Learning
Ziyan Li, Naoki Hiratani
Online learning from a stream of data is a defining feature of intelligence, yet modern machine learning systems often struggle in this setting, especially under distributional shi…
stat.ML2025
Optimal Task Order for Continual Learning of Multiple Tasks
Ziyan Li, Naoki Hiratani
Continual learning of multiple tasks remains a major challenge for neural networks. Here, we investigate how task order influences continual learning and propose a strategy for opt…
stat.ML2024
Disentangling and Mitigating the Impact of Task Similarity for Continual Learning
Naoki Hiratani
Continual learning of partially similar tasks poses a challenge for artificial neural networks, as task similarity presents both an opportunity for knowledge transfer and a risk of…