Machine Learning Starter Roadmap
Machine Learning Starter Roadmap
The plan I used to get out of tutorial hell and start building ML projects.
I spent about a year watching machine learning tutorials. I took notes, followed along and felt productive. Then I tried to build something on my own and had no idea where to start.
What fixed it was flipping the ratio: build first, and only watch a tutorial when I’m stuck. This roadmap is that approach written down, so you don’t lose the year I did.
Since then I’ve built ML models on public datasets, a RAG system over my own notes and simulations of neurons, and I share all of it on Instagram. This is where it started.
What’s inside (9 pages)
- The setup: Python, VS Code and Jupyter or Colab, done in an hour or two
- Only the Python you need before starting ML
- The core ideas to learn first: regression, classification, decision trees and scikit-learn
- The free courses I used (Kaggle, Google’s ML Crash Course) and when to use them
- 3 starter projects: Titanic survival, house prices and a spam classifier
- My “builder mode” loop and the 70/30 rule for building vs watching
- The mistakes that keep most beginners stuck
- A project tracker to keep you moving
Who it’s for
Beginners, or anyone who has watched a lot of ML content but hasn’t built anything yet. If you already build models comfortably, this will be too basic for you.
€5. Read it in 15 minutes and start your first project this week.