Course

ML for Trading: Foundations

Stefan Jansen

Author of Machine Learning for Trading

This course is popular

4 people enrolled last week.

Understand, modify, and evaluate a complete ML trading strategy

You know Python and basic machine learning. Foundations teaches you how to use them in a complete trading strategy, and how to judge the result beyond a model score.

An eight-week applied course, taken at your own pace: 42 units in eight parts, with roughly 45-55 hours of video, reading, notebooks, and exercises. Start anytime, move faster or slower, and keep lifetime access.

You work with one runnable project using 100 ETFs: prepare data, design validation, train models, turn predictions into positions, and evaluate performance after trading costs. At each stage, you modify the implementation and examine how your choices affect the result.

Recorded explanations and focused exercises give you a sequence to follow through the material. Biweekly live research hours with me let you discuss the decisions behind your results. You finish with an end-to-end implementation you can explain, change, and use as a starting point for your own research.

What you’ll learn

Learn how a trading strategy works, change its implementation, and evaluate whether the results justify further research.

  • Follow one ETF project from an economic idea through data, predictions, portfolio positions, and performance after costs.

  • Run the supplied implementation and trace how the output of each stage becomes the input to the next.

  • Connect model metrics to trading results, and explain why a better prediction does not necessarily make a better strategy.

  • Change labels, features, validation settings, and models in focused notebooks, then inspect the effect on your results.

  • Compare regularized linear models and gradient boosting on the same data and validation periods.

  • Turn predictions into positions, set portfolio constraints, and examine the trade-offs between turnover, costs, and risk.

  • Design time-aware validation and check for leakage so your model is tested on information it could actually have used.

  • Compare performance with a non-ML baseline and account for trading costs, risk, and the number of experiments you tried.

  • Explain what your backtest supports, what its assumptions leave unresolved, and what you would investigate next.

Learn directly from Stefan

Stefan Jansen

Stefan Jansen

Contact

Founder, Applied AI. Systematic investment partner since 2013. CFA.

See all products from Stefan

Who this course is for

  • Python developers and data scientists who can train a model and want to turn its predictions into a trading strategy they can evaluate.

  • Book readers and self-taught traders who want a structured sequence of explanations and exercises for applying ML to one complete project.

  • Workshop participants who have seen the full workflow and now want to understand, modify, and evaluate its individual components.

Prerequisites

  • Python with pandas/polars and NumPy

    You can read and write Python and NumPy in a notebook. The course uses Polars and introduces the operations it relies on.

  • Basic machine learning

    You have fit and scored a model with scikit-learn. Gradient boosting is taught from scratch and is not assumed.

  • No trading background required

    Markets, costs, backtests and portfolio mechanics are taught from the ground up. Prior finance experience is not needed.

What's included

Lifetime access, on your schedule

Start whenever you are ready. Return to the recorded units, notebooks, exercises, and quizzes as needed, including course updates. There are no assignment deadlines or fixed completion date.

Biweekly live research hours with Stefan

Bring questions about the code, your results, or the reasoning behind a decision. The standing Zoom sessions run every two weeks, and recordings remain available in Maven. Join when useful or catch up later.

Certificate of completion

Earn a certificate when you pass the unit checks and run the complete workflow. Your finished project gives you an implementation and results you can explain and discuss.

42 recorded units across eight parts

Work through the full sequence from strategy definition and data to validation, modeling, backtesting, portfolio construction, and evaluation. Allow roughly 45-55 hours for the videos, assigned reading, notebooks, and exercises.

Runnable notebooks and practical exercises

Start with a working ETF project. Modify its components in focused exercises, run checks on your own outputs, and see how individual choices change the result. You do not have to assemble the implementation from unrelated examples.

Course materials and prerequisite refreshers

Access videos, notebooks, quizzes, and progress tracking on ml4trading.io. Use the ML4T primer for prerequisite review. Assigned book readings are identified by section; the required book is purchased separately from the publisher.

Maven Guarantee

Your purchase is backed by the Maven Guarantee.

Course syllabus

Start here

    Start here: Welcome to Maven

    3 items

    Live research sessions: schedule and recordings

    2 items

Part 1 - Orientation and Strategy Definition

    01 - The ML4T Workflow

    1 item

    02 - The Finished Pipeline, End to End

    1 item

    03 - Trading Strategy Families and Sources of Edge

    1 item

    04 - Transaction Cost Assumptions and Performance Metrics

    1 item

    05 - Feasibility and Breadth

    1 item

Eight parts, one complete trading project

Follow the sequence below at a pace that suits you. Allow about eight weeks at 6-7 hours per week, or spread the work over longer.

1. Define the strategy. State the economic idea, trading constraints, and criteria for evaluating it.

2. Prepare data and a baseline. Build the ETF data panel and establish a non-ML comparison.

3. Design validation and labels. Specify prediction targets and test on later data without leaking future information.

4. Engineer features. Turn market data into inputs and examine the choices behind them.

5. Train and compare models. Evaluate regularized linear models and gradient boosting.

6. Backtest the strategy. Translate model predictions into simulated trades.

7. Build the portfolio. Set allocation, trading-cost, and risk assumptions.

8. Evaluate and prepare for deployment. Interpret the evidence, identify limitations, and examine how the strategy would operate.

Frequently asked questions

Maven for Teams

Reimbursement

Get your company to pay

Everything L&D needs: email template, receipts, and certificate of completion.

Get reimbursed

Private cohort

Run a cohort for your org

A dedicated cohort with a custom schedule and curriculum, tailored to your team.

Book a private cohort

$799

USD

Self-paced
Enroll