Machine Learning & Quantitative Finance

Implementation of ML/QF algorithms. Feel free to email me if you have any comment or feedback.

1. Completed

Deep Mixture Density Networks: deep MDN model implemented by Tensorflow
Dynamic Fama-French with Kalman filter: recovering latent alpha & factor loadings based on observed returns
Multi-Asset Momentum: extracting momentum signals from various asset class and test with different strategies
- 1. Adaptive weights based on signal strength (softmax activation)
- 1. Long/short mean-variance portfolio
- 1. Long/short Beta neutral portfolio
PCA Color Augmentation: a data augmentation technique widely used in image recognition
Quantile Regression: estimating q% quantile of asset return
Recommendation System:
- 1. similiarity-based recommender
- 1. collaborative filitering (Tensorflow implementation)
RNN Queue Imbalance: predicting next timestamp bid/ask direction with limit order book status, with LSTM model implemented in Keras

Fractional Differencing on Factor Returns
Python Trading Framework: an OOP framework allowed to backtest trading strategies and evaluate performance metrics

Name		Name	Last commit message	Last commit date
Latest commit History 38 Commits
Deep Mixture Density Networks		Deep Mixture Density Networks
Dynamic Fama French		Dynamic Fama French
Fractional Difference		Fractional Difference
Multi-Asset Momentum		Multi-Asset Momentum
PCA Color Augamentation		PCA Color Augamentation
Python Trading Framework		Python Trading Framework
Quantile Regression		Quantile Regression
RNN Queue Imbalance		RNN Queue Imbalance
Recommendation System		Recommendation System
README.md		README.md