paper-with-me

Papers

Exploring Feature-based Knowledge Distillation for Recommender System: A Frequency Perspective

2024-11-16 · Zhangchi Zhu, Wei zhang

In this paper, we analyze the feature-based knowledge distillation for recommendation from the frequency perspective. By defining knowledge as different frequency components of the features, we theoretically demonstrate that regular feature-based knowledge distillation is equivalent to equally minimizing losses on all knowledge and further analyze how this equal loss weight allocation method leads to important knowledge being overlooked. In light of this, we propose to emphasize important knowledge by redistributing knowledge weights. Furthermore, we propose FreqD, a lightweight knowledge reweighting method, to avoid the computational cost of calculating losses on each knowledge. Extensive experiments demonstrate that FreqD consistently and significantly outperforms state-of-the-art knowledge distillation methods for recommender systems. Our code is available at https://github.com/woriazzc/KDs.

📄 PDF Abstract BibTeX arXiv:2411.10676

Code (1)

woriazzc/kds 공식 구현 pytorch

Tasks

Knowledge DistillationRecommendation Systems

Methods 이 논문이 사용한 방법론

Knowledge Distillation A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions.…

Similar Papers 제목 키워드 기반

DE-RRD: A Knowledge Distillation Framework for Recommender System

2020-12-08 · SeongKu Kang, Junyoung Hwang, Wonbin Kweon, Hwanjo Yu

Recent recommender systems have started to employ knowledge distillation, which is a model compression technique distilling knowledge from a cumbersome model (teacher) to a compact model (student), to reduce inference la…

Knowledge DistillationModel CompressionRecommendation Systems

Rejuvenating Cross-Entropy Loss in Knowledge Distillation for Recommender Systems

2025-09-25 · Zhangchi Zhu, Wei Zhang arxiv

This paper analyzes Cross-Entropy (CE) loss in knowledge distillation (KD) for recommender systems. KD for recommender systems targets at distilling rankings, especially among items most likely to be preferred, and can o…

Knowledge Distillation

Knowledge Distillation Approaches for Accurate and Efficient Recommender System

2024-07-19 · SeongKu Kang

Despite its breakthrough in classification problems, Knowledge distillation (KD) to recommendation models and ranking problems has not been studied well in the previous literature. This dissertation is devoted to develop…

Knowledge DistillationRecommendation Systems

Preference-Consistent Knowledge Distillation for Recommender System

2023-11-08 · Zhangchi Zhu, Wei zhang

Feature-based knowledge distillation has been applied to compress modern recommendation models, usually with projectors that align student (small) recommendation models' dimensions with teacher dimensions. However, exist…

Knowledge DistillationRecommendation Systems

PromptMM: Multi-Modal Knowledge Distillation for Recommendation with Prompt-Tuning

2024-02-27 · Wei Wei, Jiabin Tang, Yangqin Jiang, Lianghao Xia 외

Multimedia online platforms (e.g., Amazon, TikTok) have greatly benefited from the incorporation of multimedia (e.g., visual, textual, and acoustic) content into their personal recommender systems. These modalities provi…

Knowledge DistillationModel CompressionRecommendation Systems