Report for Week1~2

— 杨希渊，阮墨文，陈志杰

Index

1 因子挖掘基本知识学习

•  Factor is essentially any quantity that explains or predicts the returns and risks of a group of securities

Section1: Introduction to Factor Mining

• CAPM (Capital Asset Pricing Model)
• APT (Arbitrage Pricing Theory)

Section1: Factor Models

• Preprocessing
  • Imputation & Outlier Treatment (data preprocessing)
  • Neutralization
  • Standardization
•  Single-Variable Factors & Linear Composite Factors
•  Supervised/Unsupervised Learning Approaches
  • GBDT
  • MLP
  • PCA
• Other: Genetic Algorithm & RL

Section2: Factor Construction

• Information Coefficient (IC)
  • Pearson Information Coefficient
  • Spearman Rank Information Coefficient (RIC)
• Group Backtesting Analysis
  • Assets are sorted into K groups.
  • The long-short portfolio return (factor spread).

Section3: Factor Testing

• 对于stock_data部分代码，已经跑通（包括下载代码 & 特征工程）
• 尝试使用 alphagen 及其他因子库做机器挖因子

代码部分

Thank You!