Dec 15, 2023 · To address this issue, prior studies mainly introduce item features (e.g., thumbnails) for cold-start item recommendation. They learn a feature ...
To alleviate the impact of temporal feature shifts for cold- start recommendation, we propose two new objectives for. DRO training: 1) enhancing the worst-case ...
Feb 2, 2024 · We instantiate the TDRO on two State-Of-The-Art (SOTA) cold-start recommender methods and conduct extensive experiments on three real-world ...
Linxyhaha/TDRO: Temporally and Distributionally Robust Optimization ...
github.com › Linxyhaha › TDRO
TDRO is a model-agnostic training framework and can be applied to any cold-start recommender model. You can simply create your cold-start recommender model ...
A novel temporal DRO with new optimization objectives, namely, 1) to integrate a worst-case factor to improve the worst-case performance, and 2) to devise a ...
In [55] introduced a novel approach to enhancing cold-start recommendation systems, which struggle with new items lacking historical interaction data. This ...
Feb 24, 2024 · On-demand video platform giving you access to lectures from conferences worldwide.
People also ask
What is distributionally robust optimization?
What is a robust optimization?
What is Wasserstein distance distributionally robust optimization?
2023.12: Our paper ''Temporally and Distributionally Robust Optimization for Cold-start Recommendation'' is accepted by The Annual AAAI Conference on ...
Temporally and Distributionally Robust Optimization for Cold-start Recommendation. X Lin, W Wang, J Zhao, Y Li, F Feng, TS Chua. Proceedings of the AAAI ...
This repository contains a curated list of papers on cold-start recommendations, which are sorted by their published years, hoping to provide a more ...