Stock Recommendations for Individual Investors: A Temporal Graph Network Approach with Diversification-Enhancing Contrastive Learning
By Youngbin Lee, Yejin Kim, Yongjae Lee
Rating
1494
Battle Count: 83
Relevance
8/10
Highly relevant for developing personalized stock recommendation systems that consider both user preferences and portfolio diversification
Implementation Complexity
7/10
Requires implementation of temporal graph networks and contrastive learning, but builds on existing frameworks
Reproducibility
4/5
Code and data are available on GitHub, experimental settings are well-described
About this paper
Methodology: PfoTGNRec. Problem types: Recommender Systems, Time Series Forecasting, Portfolio Optimization.
The interactive Everscope explorer (charts, battles, favorites) loads below.