Course Roadmap: machine learing/data science

Generated: 6/18/2026 | Level: Any | Format: Any

Overview

Machine learning and data science sit at the core of modern AI. Data science teaches you how to collect, clean, analyze, and visualize data, while machine learning focuses on building models that learn patterns from that data to make predictions, classifications, recommendations, and decisions. Together, they form the practical foundation for careers in analytics, AI engineering, applied machine learning, and research-oriented software development. The courses below were selected because they are real, currently available, genuinely free or free-to-audit, and come from highly reputable platforms and organizations. They also complement each other well: some build conceptual understanding, some teach hands-on Python and model building, some emphasize practical workflow and deployment, and others introduce modern deep learning and transformer-based AI. This mix gives a learner a strong path from beginner-friendly foundations to practical, portfolio-worthy skills.

Prerequisites

For the beginner courses, no major prerequisites are required beyond basic comfort using a computer. For the more technical courses, it helps to know beginner Python, high-school algebra, basic statistics, and how to work with notebooks such as Jupyter or Colab. fast.ai recommends some coding experience, preferably in Python, and Google’s ML Crash Course suggests familiarity with Python, NumPy, pandas, algebra, linear algebra, and statistics.

Estimated Time

Approximately 12-16 weeks total, or about 90-110 hours depending on pace

Learning Path

Start with AI For Everyone to build vocabulary and understand the big picture. Then take IBM: Introduction to Data Science to learn how data science work is structured in practice. After that, move into Supervised Machine Learning: Regression and Classification as your first true ML implementation course. Follow it with Google’s Machine Learning Crash Course and Microsoft Learn’s Create machine learning models to broaden your technical coverage and reinforce the full modeling workflow. Once you are comfortable with Python-based ML, take fast.ai’s Practical Deep Learning for Coders to build real deep learning projects. Finally, use the Hugging Face LLM Course to specialize in modern NLP and transformer systems after you already understand core machine learning concepts.

Recommended Courses

AI For Everyone
Provider: Coursera / DeepLearning.AI Difficulty: beginner Duration: 7 hours Free

This is a non-technical introduction to AI, machine learning, deep learning, and data science by Andrew Ng. It is ideal for complete beginners who want to understand what AI can and cannot do before diving into coding-heavy coursework.

Topics: AI fundamentals, machine learning overview, data science overview, AI strategy, ethics

Go to Course →
IBM: Introduction to Data Science
Provider: edX / IBM Difficulty: beginner Duration: 6 weeks Free

This introductory course explains what data science is, what data scientists do, and the tools, workflows, and business context behind the field. It is a strong starting point for learners who want a broad and accessible orientation before specializing.

Topics: data science foundations, data roles, tools and workflows, business applications

Go to Course →
Supervised Machine Learning: Regression and Classification
Provider: Coursera / DeepLearning.AI / Stanford Online Difficulty: beginner Duration: 3 weeks Free

This beginner-friendly Andrew Ng course teaches core machine learning ideas through linear regression, logistic regression, model training, and evaluation using Python, NumPy, and scikit-learn. It is one of the best practical first ML courses for learners who want both intuition and hands-on implementation.

Topics: supervised learning, regression, classification, scikit-learn, model evaluation

Go to Course →
Machine Learning Crash Course
Provider: Google for Developers Difficulty: intermediate Duration: 15 hours Free

Google's ML Crash Course is a fast-paced, practical program with videos, interactive visualizations, and exercises. It covers regression, classification, data handling, neural networks, overfitting, production ML, fairness, and introductory LLM concepts, making it excellent for applied learners.

Topics: regression, classification, neural networks, data preprocessing, ML fairness, production ML

Go to Course →
Create machine learning models
Provider: Microsoft Learn Difficulty: intermediate Duration: 8 hours Free

This Microsoft learning path covers core machine learning concepts alongside Python-based data exploration, regression, classification, clustering, and deep learning. It is a good structured next step for learners who want guided practice and a broader view of model types.

Topics: Python for data science, regression, classification, clustering, deep learning

Go to Course →
Practical Deep Learning for Coders
Provider: fast.ai Difficulty: intermediate Duration: 30+ hours Free

This free course is designed for people with some coding experience who want to apply deep learning to real problems. It covers computer vision, NLP, tabular data, recommendation systems, deployment, random forests, and modern PyTorch/fastai workflows, making it one of the strongest hands-on practical AI courses available.

Topics: deep learning, PyTorch, computer vision, NLP, tabular models, deployment

Go to Course →
Hugging Face LLM Course
Provider: Hugging Face Difficulty: advanced Duration: 20 hours Free

This completely free course teaches NLP and large language models using the Hugging Face ecosystem, including Transformers, Datasets, Tokenizers, Accelerate, and the Hub. It is best for learners who already know basic Python and machine learning and want to move into modern transformer-based AI.

Topics: NLP, transformers, LLMs, fine-tuning, datasets, tokenization

Go to Course →