DIFFSTOCK: PROBABILISTIC RELATIONAL STOCK MARKET PREDICTIONS USING DIFFUSION MODELS
By Divyanshu Daiya, Monika Yadav, Harshit Singh Rao
Rating
1370
Battle Count: 114
Relevance
9/10
The model shows significant improvements in stock movement prediction and portfolio management, outperforming existing methods on multiple datasets.
Implementation Complexity
8/10
The model involves complex components like diffusion models and masked relational transformers, requiring significant expertise to implement and optimize.
Reproducibility
3/5
The paper provides detailed model architecture and training procedures, but the exact implementation details and hyperparameters are not fully specified.
About this paper
Methodology: Conditional Diffusion Model with Masked Relational Transformer. Problem types: Time Series Forecasting, Classification, Portfolio Optimization.
The interactive Everscope explorer (charts, battles, favorites) loads below.