Fundamentals of Data & Python
Data Science from Scratch (Joel Grus)
Get comfortable with core Python, basic statistics, and the “why” behind data manipulation before
diving into ML libraries.
Essential Statistics for ML
Practical Statistics for Data Scientists (Peter Bruce & Andrew Bruce)
Covers the key statistical concepts (sampling, hypothesis testing, regression basics) you’ll need to
understand model behavior.
Introductory Machine Learning in Python
Introduction to Machine Learning with Python (Andreas Müller & Sarah Guido)
A gentle, code-first tour of scikit-learn’s API: preprocessing, supervised vs. unsupervised learning,
pipelines.
Deeper into Pythonic ML
Python Machine Learning (Sebastian Raschka et al.)
Builds on scikit-learn to explore more algorithms, model evaluation techniques, and begins
touching on deep learning.
Hands-On with Scikit-Learn & TensorFlow
Hands-On Machine Learning with Scikit-Learn, Keras & TensorFlow (Aurélien Géron)
Practical projects that bridge classic ML and neural-net approaches—great for consolidating
everything above.
Deep Learning Foundations
Deep Learning with Python (François Chollet)
An intuitive guide to building and training neural networks with Keras, plus clear explanations of
architectures (CNNs, RNNs).
Advanced & Specialized Topics
Feature Engineering for Machine Learning (Alice Zheng & Amanda Casari)
Techniques to transform raw data into powerful model inputs.
Probabilistic Deep Learning (Oliver Duerr et al.)
Explores Bayesian methods, uncertainty estimation, and probabilistic programming.
Reinforcement Learning (various O’Reilly titles)
When you’re ready to teach agents to make sequential decisions in environments.
Model Interpretability & Production
Interpretable Machine Learning (Christoph Molnar)
Techniques to explain “black-box” models and build trust in your predictions.
Designing Data-Intensive Applications (Martin Kleppmann)
Not strictly ML, but essential for deploying scalable, reliable pipelines in production.
Phase 1: Foundations in Python & Statistics
1. Data Science from Scratch (Joel Grus)
Learn core Python, basic linear algebra, statistics, and data-wrangling “under the hood.”
2. Practical Statistics for Data Scientists (Peter Bruce & Andrew Bruce)
Covers sampling, hypothesis testing, exploratory data analysis, regression diagnostics.
🤖 Phase 2: Core Machine Learning Algorithms
3. Introduction to Machine Learning with Python (Andreas Müller & Sarah Guido)
A hands-on dive into scikit-learn: data prep, supervised vs. unsupervised, pipelines, metrics.
4. Python Machine Learning, 3rd Ed. (Sebastian Raschka, Vahid Mirjalili…)
Builds on scikit-learn, adds ensemble methods, model selection, and an intro to neural nets.
5. Machine Learning with PyTorch and Scikit-Learn (Eli Stevens, Luca Antiga &
Thomas Viehmann)
Bridges traditional ML and deep learning, with full PyTorch workflows.
⚙️Phase 3: Comprehensive Hands-On Projects
6. Hands-On Machine Learning with Scikit-Learn, Keras & TensorFlow (Aurélien Géron)
End-to-end projects: from classic algorithms through CNNs/RNNs, with production tips.
7. Deep Learning with Python (François Chollet)
Keras-first deep architectures (CNNs, RNNs, autoencoders) and best practices for real tasks.
8. Generative Deep Learning (David Foster)
Dive into VAEs, GANs and autoregressive models to build creative AI applications.
🔍 Phase 4: Feature Engineering & Interpretability
9. Feature Engineering for Machine Learning (Alice Zheng & Amanda Casari)
Systematic approaches to cleaning, transforming and selecting features for all data types.
10.Interpretable Machine Learning (Christoph Molnar)
Techniques (SHAP, LIME, partial dependence) to explain complex “black-box” models.
Phase 5: Architecture & Patterns
11.Machine Learning Design Patterns (Valliappa Lakshmanan, Sara Robinson &
Michael Munn)
A cookbook of repeatable solutions: data ingestion, batch vs. streaming inference, A/B
testing, etc.
12.Designing Machine Learning Systems (Chip Huyền)
Holistic systems thinking: data ownership, retraining schedules, monitoring and feedback
loops.
13.AI Engineering (Chip Huyền)
How to select foundation models, define benchmarks, and deploy scalable AI services.
🚀 Phase 6: Production & MLOps
14.MLOps: Continuous Delivery and Automation for Machine Learning (Mark Treveil &
Alok Shukla)
CI/CD for data and models, dataset versioning, automated testing, canary rollouts.
15.Kubeflow for Machine Learning: From Lab to Production (Trevor Grant et al.)
Build, tune (Katib) and serve models at scale on Kubernetes with Kubeflow pipelines.
16.Designing Data-Intensive Applications (Martin Kleppmann)
Fundamental patterns for storage engines, streaming vs. batch, consistency and
fault-tolerance.
17.Observability Engineering (Charity Majors, Liz Fong-Jones & George Miranda)
Master logs, metrics and tracing for reliable, performant ML services.
📈 Phase 7: Advanced & Probabilistic Methods
18.Bayesian Methods for Hackers (Cam Davidson-Pilon)
Intuitive, code-first intro to Bayesian inference with PyMC3.
19.Probabilistic Deep Learning (Oliver Duerr, Andreas Damgaard & Søren Hauberg)
Marry deep nets with probabilistic uncertainty estimation and Bayesian neural networks.
20.Reinforcement Learning with Python (Abhishek Nijhara & Shalabh)
Hands-on guide to teaching agents with OpenAI Gym, policy gradients and value-based
methods.
📚 How to Use This Roadmap
• Build phase by phase: master each level before moving on.
• Code as you read: clone the GitHub repos or retype examples.
• Small projects: pick a public dataset or simple toy problem for every book.
• Iterate & revisit: come back to earlier phases after you’ve seen advanced concepts.
This sequence will take you from zero ML experience through state-of-the-art, production-grade
systems—using every major O’Reilly title along the way. Enjoy the journey!
• AUTOMATE BORING STUFF WITH PYTHON
• os & API
