KittenCN/stock_prediction

基于神经网络的通用股票预测模型 A general stock prediction model based on neural networks

48
/ 100
Emerging

Combines multi-scale architectures (TemporalHybridNet, PTFT+Variational SSM, Diffusion, Graph) with a complete data pipeline supporting tushare/akshare/yfinance ingestion, automatic feature engineering (log returns, macro/sentiment exogenous variables, stock embeddings), and per-symbol normalization. Provides unified PyTorch training with 10+ model variants, automatic evaluation metrics (RMSE, MAPE, quantile coverage, VaR), and consistent inference workflows with serialized normalization parameters.

357 stars.

No Package No Dependents
Maintenance 6 / 25
Adoption 10 / 25
Maturity 9 / 25
Community 23 / 25

How are scores calculated?

Stars

357

Forks

86

Language

Python

License

GPL-3.0

Last pushed

Oct 20, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/KittenCN/stock_prediction"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.