How to use the Machine Learning Roadmap
- 1. Tick what you knowOpen each stage and tick the topics you can already do confidently.
- 2. Pick your paceChoose how many hours a week you can study; the estimate updates for what's left.
- 3. Get your planSee time left per stage and your next three topics, and download the checklist.
Who this roadmap is for
This roadmap is for students and developers who want to build and ship machine learning models. It puts classic ML before deep learning on purpose: tree models and logistic regression are often the right answer for tabular business data, and they teach evaluation habits that neural networks also need.
The maths stage covers only what you'll use: vectors, gradients, probability and basic statistics.
The 8 stages
Each of the 8 stages of the Machine Learning Engineer path ends with a milestone project, because building something is what turns topics into skills.
- Python for ML (about 3 weeks): Python fundamentals and virtual environments, NumPy arrays and vectorisation, pandas for loading and cleaning data and more. Milestone: Clean and explore a public dataset in a notebook.
- Maths you need (about 4 weeks): Vectors, matrices and dot products, Derivatives, gradients and gradient descent, Probability, distributions and Bayes' rule and more. Milestone: Implement linear regression with gradient descent in NumPy.
- Classic machine learning (about 6 weeks): Supervised vs unsupervised learning, Linear and logistic regression, Decision trees, random forests, gradient boosting and more. Milestone: Predict an outcome on tabular data and beat a simple baseline.
- Evaluation (about 2 weeks): Train/validation/test splits and cross-validation, Precision, recall, F1 and ROC AUC, MAE, RMSE and R² and more. Milestone: Write a model report comparing three models fairly.
- Deep learning (about 6 weeks): Neural networks and backpropagation, PyTorch tensors, autograd and training loops, CNNs for images and more. Milestone: Image classifier with transfer learning.
- NLP and LLMs (about 4 weeks): Tokenisation and embeddings, Transformers in plain words, Using pre-trained models from Hugging Face and more. Milestone: A question-answering app over your own documents.
- Deploying models (about 3 weeks): Saving models and versioning, Serving with FastAPI, Docker for ML apps and more. Milestone: Deploy one model as an API with a simple web front end.
- Job ready (about 3 weeks): Portfolio: 3 end-to-end projects with READMEs, Explain model choices and trade-offs, ML interview topics and case questions and more. Milestone: Publish a write-up of your best project's results and limits.
How the time estimate works
Worked example: with nothing ticked and 10 hours a week, the plan shows about 31 weeks for all 36 topics. Choose 20 hours a week and it becomes about 15.5 weeks. Tick every topic in the first stage, Python for ML, and that row shows "done" while the total at 10 hours a week drops by 3 weeks to about 28.
Week counts are our suggestion for steady self-study of the Machine Learning Engineer path, based on how much each stage covers. They are not survey data or a promise; use them to plan, not to judge yourself.
Tips for following it
- Implement one algorithm from scratch (linear regression is ideal) before relying on libraries.
- Always compare against a simple baseline; a model that can't beat it isn't ready.
- Write down your data split before training and never tune on the test set.
Frequently asked questions
Machine learning or data science roadmap?
The data science roadmap leans towards analysis and business questions. This one leans towards building, deploying and maintaining models.
Do I need a GPU?
Not for classic ML. For deep learning, free notebook GPUs are enough for the milestone projects.
Where are my ticks saved?
Your Machine Learning Roadmap ticks are kept only in this browser on this device, so nothing you tick is sent anywhere. Clearing site data or switching browsers starts you fresh; download your checklist to keep a copy.
Is the Machine Learning Roadmap free?
Yes. Reading the Machine Learning Roadmap and ticking topics is open to everyone. Viewing your plan the first time asks for your name and WhatsApp number once; after that every Uncodemy tool is free and open to you.