Specialist track, course code ML
Machine Learning for Finance
Machine learning in finance, applied with the controls that markets demand: purged validation, leakage checks and honest costs.
For machine learning in finance, the data is the hard part. Signals are weak, samples are small relative to the noise, observations overlap in time, and the process changes as other traders learn the same patterns. Methods that work on images and text fail here unless validation is designed for these conditions.
What does the machine learning track cover?
A free course reviews what machine learning has and has not achieved in trading. Machine Learning for Trading covers feature construction, tree ensembles and walk-forward validation on market data. Financial Machine Learning, at Researcher level, covers purged and embargoed cross-validation, meta-labelling, sample weighting and backtest overfitting. Reinforcement Learning for Trading covers policy learning for execution and allocation and why reward design matters.
Where does this track lead?
To the Certificate in Machine Learning for Trading, which ends in a capstone model validated out of sample with costs charged.
Open courses in this track
Free primers. Read, watch and work the exercises before you spend anything.
Practitioner courses in this track
Lab-based courses. You build, test and cost real strategies and systems on market data.
Researcher courses in this track
Proof-level courses. Stochastic models, estimation theory and the literature behind each method.
- Financial Machine Learning The research methods developed for machine learning on financial data: sampling with information-driven bars, triple-barrier labels and meta-labeling, sample weights for overlapping outcomes, purged cross-validation, feature importance, and statistics of backtest overfitting.
- Reinforcement Learning for Trading Reinforcement learning applied to sequential trading decisions: Markov decision processes, dynamic programming, Q-learning and policy gradient methods, and their use in optimal execution, market making and portfolio allocation, with the simulation and evaluation problems that limit them.
Credentials that use this track
- Certificate in Machine Learning for TradingFive courses and a capstone on applying machine learning to markets without fooling yourself: labelling, features, purged validation, overfitting tests and reinforcement learning.
- Fellowship in Computational FinanceThe top credential on the researcher path. Nine courses, eight of them at researcher level, and a supervised dissertation, examined by two faculty members.
Questions applicants ask
Does AI trading work?
Some machine learning methods add value in specific tasks such as execution, risk forecasting and combining many weak signals. Most published results on price prediction do not survive costs and proper validation. The free course AI Trading: What Works and What Does Not covers the evidence.
What background do I need?
Python and statistics. The Certificate in Machine Learning for Trading includes Python for Finance and Time Series Analysis for that reason.