Saved in:
Bibliographic Details
Main Authors: Li, Dong, Jin, Ruoming, Ren, Bin
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2312.08520
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916465257480192
author Li, Dong
Jin, Ruoming
Ren, Bin
author_facet Li, Dong
Jin, Ruoming
Ren, Bin
contents Inspired by the success of contrastive learning, we systematically examine recommendation losses, including listwise (softmax), pairwise (BPR), and pointwise (MSE and CCL) losses. In this endeavor, we introduce InfoNCE+, an optimized generalization of InfoNCE with balance coefficients, and highlight its performance advantages, particularly when aligned with our new decoupled contrastive loss, MINE+. We also leverage debiased InfoNCE to debias pointwise recommendation loss (CCL) as Debiased CCL. Interestingly, our analysis reveals that linear models like iALS and EASE are inherently debiased. Empirical results demonstrates the effectiveness of MINE+ and Debiased-CCL.
format Preprint
id arxiv_https___arxiv_org_abs_2312_08520
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Revisiting Recommendation Loss Functions through Contrastive Learning (Technical Report)
Li, Dong
Jin, Ruoming
Ren, Bin
Artificial Intelligence
Inspired by the success of contrastive learning, we systematically examine recommendation losses, including listwise (softmax), pairwise (BPR), and pointwise (MSE and CCL) losses. In this endeavor, we introduce InfoNCE+, an optimized generalization of InfoNCE with balance coefficients, and highlight its performance advantages, particularly when aligned with our new decoupled contrastive loss, MINE+. We also leverage debiased InfoNCE to debias pointwise recommendation loss (CCL) as Debiased CCL. Interestingly, our analysis reveals that linear models like iALS and EASE are inherently debiased. Empirical results demonstrates the effectiveness of MINE+ and Debiased-CCL.
title Revisiting Recommendation Loss Functions through Contrastive Learning (Technical Report)
topic Artificial Intelligence
url https://arxiv.org/abs/2312.08520