C
Chenxu Wang
Researcher at University of Science and Technology of China
Publications - 3
Citations - 78
Chenxu Wang is an academic researcher from University of Science and Technology of China. The author has contributed to research in topics: Overfitting & Meta learning (computer science). The author has an hindex of 1, co-authored 2 publications receiving 34 citations.
Papers
More filters
Proceedings ArticleDOI
How to Retrain Recommender System?: A Sequential Meta-Learning Method
TL;DR: This work proposes a new training method, aiming to abandon the historical data during retraining through learning to transfer the past training experience, and designs a neural network-based transfer component, which transforms the old model to a new model that is tailored for future recommendations.
Proceedings ArticleDOI
How to Retrain Recommender System? A Sequential Meta-Learning Method
TL;DR: In this paper, a neural network-based transfer component is designed to transform the old model to a new model that is tailored for future recommendations, which can account for both long-term and short-term user preference.
Journal ArticleDOI
Deep Reinforcement Learning with Multitask Episodic Memory Based on Task-Conditioned Hypernetwork
TL;DR: In this article , a retrieval network based on a task-conditioned hypernetwork is proposed, which adapts the retrieval network's parameters depending on the task and enhances the collaborative efforts between the retrieval and decision networks.