Machine learning foundations

Junior data analyst, ML beginner, technical student — I train a model, measure it honestly and know when not to use it.

Level 2 — technical (Python)

You can already write Python and handle a dataframe. Now you want to know what a model is, how it is trained, and why most of the work is not the model.

What you'll be able to do

• Frame a problem as regression or classification, or decide it is neither
• Split data properly and explain why a test set exists
• Train, tune and compare models with scikit-learn
• Read accuracy, precision, recall and a confusion matrix without fooling yourself
• Detect overfitting and leakage in your own notebook
• Refuse a use case where the model would cause harm

Modules

• M1 — How a machine learns — Features, labels, fitting and the error being minimised.
• M2 — Training and testing — Splits, cross-validation, and why your first score was a lie.
• M3 — Classification — Yes/no problems and the metrics that matter.
• M4 — Feature work — Encoding, scaling and the features you create yourself.
• M5 — Harm, bias and refusal — Credit, hiring, health: where a model must not decide alone.

Capstone project

A model, measured honestly — Take a real problem, build a baseline, train at least two models, evaluate them on unseen data, compare per group, and write what you would tell a manager — including whether to deploy it at all.

What you earn

• A trained model with a written evaluation
• A Kaggle profile with submitted work
• LES Institute certificate — Machine Learning Foundations

Requirements

• Python for Data, or equivalent Python and pandas experience
• Secondary-school mathematics, including basic statistics