Machine LearningFoundationOne To One
Machine Learning from First Principles
Build, evaluate and explain models on real data, one algorithm at a time.
About this programme
Twelve sessions that take you from a raw dataset to a model you can defend. Each session covers one idea, then you implement it in Python on a dataset you choose, and explain the result back before we move on. This follows the machine learning course I teach to B.Tech students at Bennett University, with more time for practice and for your own questions.
What you will be able to do
- Prepare a raw dataset for modelling
- Train and tune regression and classification models
- Choose the right metric and validate without leaking data
- Explain what a model learned and where it fails
Curriculum
4 modules · 12 lessons
- What a model learns and how we measure it1h
- Cleaning data and handling missing values1h
- Features, scaling and encoding1h
Before you start
- Python basics: lists, functions and loops
- School level algebra and some statistics