Learning paths

Every module can be entered on its own, but a path gives you an order and a reason for it. Each path keeps the course's core habit: decide what evidence a result needs before you trust it. Time estimates on each module page are the author's estimates; measured learner times will replace them after piloting.

Practical foundations: new to programming or machine learning

1. Take the readiness check; use Python Foundation and the Python-to-ML bridge where it suggests.
2. Work through Modules 1 to 4: rules versus models, files and functions, cleaning, exploring.
3. Continue to Module 5 (your first trained model and baseline) and Module 6 (classification and thresholds).
4. Finish Module 6 with the stability exercise.

At every module, answer the "Check yourself" questions on the concepts page before practice, and use the study workflow for what to keep.

IT operations analyst: applying ML to operational data

1. Modules 2, 3 and 4: tickets, network events and service workload.
2. Module 6, then the historical phishing case and the RT-IoT intrusion triage transfer.
3. Module 12: monitoring, then the JSON API extension.
4. Module 10's retrieval extension and Module 14: grounding answers and bounding tools.

Evaluation-focused practitioner: you already build models

1. Skim the readiness check, then Module 5 for the evaluation protocol the course uses.
2. In Module 6, run the stability, tuning and encoding extensions (the encoding one is in Module 3).
3. The complete capstone, the advanced investigations and the research comparison guide.
4. Modules 8 to 11 for neural, language-model and fine-tuning evaluation.

The recurring IT case thread

IT examples recur through the course alongside other datasets. Each has limits, and stating them is part of using it well.