Getting Started

Installation
Video Lectures
Text Books
QuantConnect
Special Mention: Prof M Lopez de Prado

API Reference

mlfinlab
- Subpackages

Backtest Overfitting Tools

7 Reasons Most ML Funds Fail
7 Point Backtesting Protocol
Backtest Statistics

Data Structures

Data Preparation Tutorial
Single Futures Roll
Standard Bars
Imbalance Bars
Run Bars
Online Data Structures

Labeling

Introduction To Labeling
Triple-Barrier
Meta-Labeling
Trend Scanning
Raw Returns
Fixed Horizon
Excess Over Mean
Excess Over Median
Return Versus Benchmark
Bull Bear States
Matrix Flags
Tail Sets

Sampling

Introduction to Sampling
Sample Uniqueness
Sequential Bootstrapping
Sample Weights

Feature Engineering

Fractionally Differentiated Features
Structural Breaks
Filters
Volatility Estimators
Noise Reduction
Directional Change

Microstructural Features

First Generation Models
Second Generation Models
Third Generation Models
Entropy Measures
Message Encoding

Models

Sequentially Bootstrapped Ensembles
History Weighted Regression

Clustering

Optimal Number of Clusters (ONC)
Hierarchical Clustering

Cross-Validation

Purged and Embargo
Combinatorial Purged CV (CPCV)

Feature Importance

Model Fingerprint Algorithm
MDI, MDA, and SFI
Clustered MDA and MDI
Shapley Values
Feature Clustering
PCA Features and Analysis

Bet Sizing

Bet Sizing in ML
Exact Fit using first 3 Moments (EF3M)
Kelly Criterion

Synthetic Data Generation

Introduction
CorrGAN
Vine and Extended Onion Methods
Correlated Random Walks
Hierarchical Correlation Block Model (HCBM)
Bootstrapping
Data Verification

Networks

Introduction
Minimum Spanning Tree (MST)
Average Linkage Minimum Spanning Tree (ALMST)
Planar Maximally Filtered Graph (PMFG)
Visualising Graphs
Dash Network Visualisations

Measures of Codependence

Introduction
Correlation-Based Metrics
Information Theory Metrics
Copula-Based Metrics
Optimal Transport
Codependence Matrix

Additional Resources

Research Tools
Sample Datasets
Barriers to Entry

Developer Guide

Documentation for Debugging
Changelog

mlfinlab

`mlfinlab.regression`

Implementation of historically weighted regression method based on relevance.

Submodules

mlfinlab.regression.history_weight_regression

Previous Next