ML for Trading: From Research to Production
Stefan Jansen
ML for Trading · Applied AI · Since 2013
Learn to make the decisions rigorous trading research requires.
A trading strategy is not a model followed by a backtest. It is a chain of decisions about the question, market, data, timing, labels, features, validation, models, portfolio, costs, risk, and production. Weak research often fails at the connections between those decisions, even when the code runs and the score looks attractive.
Research to Production teaches you how to conduct that work. You get a working codebase and nine case studies spanning ETFs, equities, futures, FX, crypto, and options. Extend the pipeline, point it at a different market, or apply the workflow to your own strategy. Learn how to interpret the evidence, identify what deserves further work, and decide what to do next.
By the end, you will have a cumulative research record that makes your reasoning and evidence intelligible: what you examined, what you reproduced or changed, what the results support, what remains uncertain, and whether the strategy should proceed, be revised, be monitored, or stop. I will review that work with you in a 30-minute one-on-one.
What you’ll learn
Conduct rigorous, iterative research across the full trading pipeline, with a working codebase and nine case studies in different markets.
Connect a market mechanism to a tradable universe, decision clock, holding period, horizon, and feasibility constraints.
State the evidence that could change your view before selecting data or a model.
Compare how the same decision plays out across nine markets, with the evidence prepared for you.
Make timing, identity, availability, aggregation, and information boundaries explicit before interpreting a model.
Align labels and features with the tradable action, then fit every learned transformation inside the temporal folds.
Compare construction choices across market, fundamental, alternative, and multi-asset data.
Use walk-forward validation, protected holdouts, fold-level evidence, and search accounting to compare like with like.
Ask whether boosting, deep learning, latent factors, or causal methods answer the question better than a simple reference.
Interpret attribution and uncertainty without mistaking either for causality.
Specify the backtest and portfolio, then test turnover, concentration, costs, execution, risk, capacity, and robustness.
Separate value created by the signal from value created or destroyed by position mapping and allocation.
Assess research, shadow, paper, and live readiness as distinct stages.
Reconstruct and audit a case, modify and extend it, or adapt the workflow to a new application.
Choose deeper predictive, structural, or causal work only when the research question warrants it.
Use core-path guidance and optional practice without imposing a uniform number of models or iterations.
Synthesize what the evidence supports, what remains uncertain, and what deserves further investigation.
Distinguish research, economic, risk, and production readiness.
Conclude with a proceed, revise, investigate, monitor, or stop decision.
Learn directly from Stefan
Stefan Jansen
Author, ML for Trading (3rd ed) · 9 case studies · Applied AI
Who this course is for
Data scientist, ML engineer, quant researcher, or trader seeking a rigorous workflow across data, models, strategy design, and production.
Investment professional or technically strong practitioner who knows one part of the process and wants better end-to-end research decisions.
Transition candidate with Python and basic ML who wants a demanding applied course, not a casual introduction or ready-made strategy.
Prerequisites
Python fluency
Comfortable writing Python, working with pandas and NumPy, and reading a moderately complex research codebase.
Basic statistics and ML
You have trained and evaluated a supervised model at least once. The course reviews methods; it does not teach them from zero.
Serious interest in systematic research
You want to reason about market data, validation, and trading decisions, not run a fixed recipe.
What's included
Live sessions
Learn directly from Stefan Jansen in a real-time, interactive format.
15 hours live with Stefan
Ten 90-minute live sessions, one in each of the first ten weeks. Between sessions, Stefan answers questions in the cohort's #questions channel.
30-minute individual research review
Offered to every student and scheduled within a month of the final session. Discuss your research record, your conclusions, and the most valuable next step directly with Stefan.
A working pipeline and nine case studies
Extend the supplied pipeline, point it at another market, or bring your own, and compare your decisions against worked research in ETFs, equities, futures, FX, crypto, and options.
Continued access to course materials
Return to the recordings, handouts, optional exercises, book references, and course repository after the cohort ends.
Weekly handouts and optional practice
Follow a core project path and choose additional exercises that match your background, research direction, and available time.
112-topic ML4T primer
Use concise, publish-ready topic packets for prerequisite review, technical refreshers, and deeper study throughout the course.
Maven Guarantee
Your purchase is backed by the Maven Guarantee.
Course syllabus
10 live sessions • 54 lessons
Week 1
01 - Your research environment
02 - The study folder, the findings ledger, and the shared protocol
03 - The nine case studies and their trading ideas
04 - Using coding agents in this course
Sep
16
Live session: Orientation, nine case studies, and your first runs
Week 2
01 - Asset class, instrument, vehicle, and target capital
02 - Bars: time, tick, volume, dollar, imbalance
03 - Reference price, roll rules, and back-adjustment
04 - Point-in-time correctness under back-adjustment
05 - Jumps: remove, flag, or keep as signal
06 - Entity matching and whether alternative data earns its burden
Sep
23
Live session: Data Infrastructure
Schedule
Live sessions
2 hrs / week
A 90-minute session in each of the first ten weeks: the week's decision answered on a second market, an experiment run on the shared pipeline, and next week's experiment worked out live.
Wed, Sep 16
4:00 PM—5:30 PM (UTC)
Wed, Sep 23
4:00 PM—5:30 PM (UTC)
Wed, Sep 30
4:00 PM—5:30 PM (UTC)
Independent research
3-4 hrs / week
Advance a cumulative research record at your own depth: reconstruct and audit a case, modify or extend it, or adapt the workflow to your application. This is the core tier; optional extensions and stretch work sit above it, and about one student in ten takes them.
Reading and optional practice
1-2 hrs / week
Use the weekly handout to select book sections, notebooks, and optional exercises that support your current research question and learning goals.
Frequently asked questions
Maven for Teams
Reimbursement
Get your company to pay
Everything L&D needs: email template, receipts, and certificate of completion.
Get reimbursedPrivate cohort
Run a cohort for your org
A dedicated cohort with a custom schedule and curriculum, tailored to your team.
Book a private cohort