Quantitative Trading with Machine Learning
summary
Company Quantiacs presents hands-on results using MATLAB and Python toolboxes to build a neural network auto-regressive model that predicts next-day futures returns after a 75/25 in/out split. Simple models show in-sample correlation ~30% but test collapse to ~1%, expanding lookback improves to ~12% in-sample and ~4.5% test yet still noisy, so preprocessing, custom objective functions and longer horizons matter. A competition-winning system is live trading $1M futures with a Sharpe ratio of 2, and Quantiacs takes a 10% cut while investors take 80%.
AI-written summary · about
chapters (15)
- 0:00Intro
- 0:43Quantitative Finance
- 2:571st Marketplace for trading algorithms
- 4:32Machine learning for quantitative trading
- 9:46Predicting a time series
- 11:28Building a model in Matlab
- 23:58Build a model
- 25:25Preprocess the data
- 28:09Define y(t)
- 31:16Objective function
- 32:31Sharpe Ratio
- 36:42Good practice and pitfalls
- 38:47Quantiacs' market data
- 40:08Loading the market data
- 45:49galvanize
related talks
33mQuantamental Investing - the future of finance
07mQ n A with Dr Tom Starke
30mHow to build a stock market money printing machine with AI
23mComplex AI Forecasting Methods for Investments
27mModelling the extreme using Quantile Regression
08mQ n A with Dr Arun Verma
30mPredictive trading
21mOpportunities for Quantum Computing in Finance