Boosting: Learning Slowly on Residuals — MatheLinux

MatheLinuxQuant learning system

Boosting: Learning Slowly on Residuals — MatheLinux

AI & Machine Learning for Finance

0%

0 of 50 lessons complete

Layer 1Preview layerFree▾

Layer 2Statistical Learning FoundationsPro▾

The Bias-Variance Tradeoff8m
Assessing Model Accuracy8m
Framing Quant Problems as Supervised Learning8m

Layer 3Linear Models & Factor ModelsPro▾

Simple Linear Regression and Least Squares8m
Multiple Regression in Matrix Form8m
Alpha and Beta: The CAPM Regression8m
Factor Models and Idiosyncratic Risk8m

Layer 4Classification & Credit DefaultPro▾

Logistic Regression and the Probability of Default8m
LDA, QDA, and Naive Bayes8m
Structural Credit Models: Distance to Default and Moody's EDF8m
Evaluating Default Classifiers: ROC, AUC, and Thresholds8m

Layer 5Validation & Backtest OverfittingPro▾

Cross-Validation: Estimating Test Error Honestly8m
The Bootstrap and Its Limits in Finance8m
Subset Selection, Ridge, and the Lasso8m
Backtest Overfitting and the Deflated Sharpe Ratio8m

Layer 6Splines, Trees & BoostingPro▾

Beyond Straight Lines: Splines and GAMs8m
Decision Trees: Greedy Splits and Their Instability8m
Bagging and Random Forests: Averaging Away Variance8m
Boosting: Learning Slowly on Residuals8m

Layer 7Support Vector MachinesPro▾

Separating Hyperplanes and the Maximal Margin8m
Soft Margins: The Support Vector Classifier8m
The Kernel Trick: Nonlinear Boundaries from Inner Products8m
SVMs on the Trading Desk: Direction, Credit, and Limits8m

Layer 8Deep Learning for FinancePro▾

Neural Networks: Anatomy of a Single Layer8m
Training Networks: Gradient Descent and Backpropagation8m
Regularizing Networks: Dropout, Early Stopping, Weight Decay8m
Sequence Models: RNNs, Attention, and Deep Hedging8m

Layer 9Survival Analysis & Time-to-DefaultPro▾

Censored Data, Survival and Hazard Functions8m
The Kaplan-Meier Estimator and the Log-Rank Test8m
Cox Proportional Hazards and Hazard Ratios8m
Modeling Time to Default: Survival in Production Credit8m

Layer 10Unsupervised Learning & Market RegimesPro▾

PCA: Level, Slope, Curvature8m
Clustering Assets and Trading Days8m
Detecting Market Regimes8m
Eigen-Portfolios and Denoising Risk8m

Layer 11Multiple Testing, the Factor Zoo & LLMsPro▾

The Multiple Testing Problem8m
The Factor Zoo and P-Hacking8m
Attention: Queries, Keys, Values8m
LLMs in Finance: Uses and Limits8m

Layer 12FoundationsPro▾

Linear Regression & Alpha/Beta8m
Classification & Credit Risk8m

Layer 13Validation & Flexible ModelsPro▾

Resampling & Model Selection8m
Moving Beyond Linearity8m

Layer 14Ensembles & Modern AIPro▾

Tree-Based Methods8m
Support Vector Machines8m
Deep Learning & Transformers in Finance8m

Layer 15Unsupervised & Statistical RigorPro▾

Unsupervised Learning8m
Survival Analysis & Multiple Testing8m