loss function

3 papers across 2 sessions

Poster Session 4

Thursday, December 4, 2025 · 4:30 PM → 7:30 PM

Graph-Theoretic Insights into Bayesian Personalized Ranking for Recommendation

#3813 · Kai Zheng, Jianxin Wang, Jinhui Xu

This paper introduces BPR+, a novel loss function for graph self-supervised learning that extends Bayesian Personalized Ranking (BPR) by incorporating even-hop paths to better capture global connectivity and topological structure.

Any-stepsize Gradient Descent for Separable Data under Fenchel–Young Losses

#3007 Spotlight · Han Bao, Shinsaku Sakaue, Yuki Takezawa

Poster Session 6

1 paper

Friday, December 5, 2025 · 4:30 PM → 7:30 PM

Exhibit Hall C,D,E

DBLoss: Decomposition-based Loss Function for Time Series Forecasting

#2302 · Xiangfei Qiu, Xingjian Wu, Hanyin Cheng, Xvyuan Liu, Chenjuan Guo, Jilin Hu, Bin Yang