Where to Learn Data Science on YouTube, by Skill Level

If you want to learn data science on YouTube, start with freeCodeCamp.org or Programming with Mosh for Python, use StatQuest for statistics and machine learning theory, and switch to Alex The Analyst or Luke Barousse for SQL, Excel and portfolio projects. Then add Krish Naik or DeepLearning.AI once you are ready for machine learning and AI.

Data science changes fast, and YouTube plays an incredible role because the best teachers there are free and keep publishing. Paid “gurus” used to be the only option. Today you can build most of a self-study roadmap from the channels below.

We checked every channel in October 2026 to confirm what it covers now and whether it still uploads. A few have changed focus or slowed down, and we say so in each entry.

Quick Picks by Goal

Krish Naik

Best for: learners who want one channel that covers the whole path, from Python and statistics to deep learning, NLP and MLOps.

Krish Naik’s channel is a large library for aspiring data scientists. His playlists follow the data science lifecycle step by step, and his end-to-end project tutorials show how the pieces fit together in a real workflow. In 2026 he also posts a lot about AI jobs and which roles to prepare for, which is useful if you are planning a career switch.

StatQuest with Josh Starmer

Best for: anyone who finds the math and statistics behind data science intimidating.

Josh Starmer breaks down statistical concepts and machine learning algorithms into simple, memorable steps, with his trademark jingles and drawings. You won’t just learn how an algorithm works; you’ll understand why. He still publishes regularly, with recent videos on Naive Bayes and the activation functions used in modern neural networks. If a concept from another course didn’t click, search for it here first.

freeCodeCamp.org

Best for: people who prefer one long, structured course over many short videos.

This nonprofit channel publishes full-length, university-style courses for free, and data science is well covered. You can find multi-hour tutorials on Python for data analysis, complete machine learning courses, and deep dives into libraries like Pandas and TensorFlow. New courses still go up several times a week, and many of them now cover AI engineering topics as well.

Alex The Analyst

Best for: the data analyst track, especially if you want a job rather than a research career.

Alex Freberg focuses on the core skills analysts use every day: SQL, Excel, Tableau and Power BI. He walks viewers through portfolio projects from scratch and gives a clear roadmap from learning to landing a job. He is currently running an intro to statistics series, which is a good starting point if you skipped statistics in school.

Luke Barousse

Best for: analysts who want a free, bootcamp-style course and people curious about data engineering.

Luke Barousse publishes long, free courses built around real projects. In 2026 he released a 35+ hour Data Analyst Bootcamp (SQL, Excel, Power BI, Python and GitHub), a 27+ hour Data Engineer Bootcamp (SQL, Python, cloud, Bash and Git), and a full dbt course. If you searched for data analytics or data engineering channels, start here.

Ken Jee

Best for: career advice from his older data science videos.

Ken Jee built his audience with career advice on portfolios, interviews and breaking into data science, plus the #66DaysOfData learning challenge. His data science tutorials and career videos remain worth watching as an archive. Note that the channel has changed direction: since 2024 he has mostly posted about preparing for AI disruption and building his own software products, so don’t expect new data science tutorials.

Programming with Mosh

Best for: absolute beginners who need to learn Python before anything else.

Mosh Hamedani’s channel isn’t data-science-specific, but his Python tutorials are known for their clarity, production quality and structure. A strong command of Python is essential for data science, and his beginner Python course is a good first step. Uploads have slowed (the latest, in March 2026, was a Claude Code tutorial), but the existing courses don’t go out of date quickly.

365 Data Science

Best for: visual learners who want short, animated theory lessons.

365 Data Science built a library of professionally produced, animated videos on statistics, probability, machine learning fundamentals and data literacy. Treat it as an archive: in 2026 the channel has mostly posted student testimonials and promotions for its paid platform, and its last educational uploads (a portfolio-building guide and ChatGPT explainers) date from late 2025.

Tina Huang

Best for: self-learners who want a study strategy and an honest view of tech careers.

Tina Huang is an ex-Meta data scientist. Her channel now focuses on AI, coding, tech careers and self-study, with videos such as AI roadmaps and quick walkthroughs of tools. Watch her if you need help deciding what to learn next and how to study efficiently, rather than for step-by-step data science lessons.

Corey Schafer

Best for: learners who already know basic Python and want to write cleaner, more professional code.

Corey Schafer’s tutorials are detailed and clear. His playlists on Pandas, Matplotlib and object-oriented programming are still some of the best free resources for the programming side of data work. He posts occasionally; in 2026 he released a full FastAPI course, which is handy once you want to serve a model or a dashboard through an API.

DeepLearning AI

Best for: learners who have the basics and want to go deeper into machine learning and artificial intelligence.

Founded by AI pioneer Andrew Ng, DeepLearning.AI posts videos from its machine learning courses, recorded events and talks with AI researchers and engineers. Recent uploads cover AI agents and practical AI engineering topics. It’s a good next step once you are comfortable with Python and core ML concepts.

Looking for Coding Channels Instead?

Many people who search for data science channels really want general coding channels. freeCodeCamp.org, Programming with Mosh and Corey Schafer (above) are good for that too. A few more that were still posting in October 2026:

• CS50: free computer science and math courses from the team behind Harvard’s CS50, led by David J. Malan. A good choice if you want fundamentals before choosing a specialty.
• Fireship: fast, short code tutorials and developer news. Good for staying current, not for learning from zero.
• Traversy Media: practical web development and programming tutorials, from HTML, CSS and JavaScript to React, Node.js and Python.
• Web Dev Simplified: clear explanations of JavaScript and web development concepts.
• Net Ninja: long, step-by-step playlists on web frameworks and developer tools.

If you’d rather listen while commuting, see our list of coding podcasts for developers.

Next Steps

You’ve checked out the list, subscribed to a few creators, and maybe even turned on notifications.

But watching videos alone won’t make you a data scientist. Too many tabs and half-finished playlists can bury you in noise. The habit that matters is to take notes from every lesson, then write down one thing to practise in code before the next video.

An AI summarizer can help with the note-taking part. With TubeOnAI you can summarize long YouTube tutorials, ask follow-up questions about a summary, and save summaries in Collections so your notes stay organized by topic. You can also subscribe to YouTube channels so new videos are summarized automatically, which helps you skim what’s new on fast-moving channels before deciding what to watch in full.

If you’re interested, get started with TubeOnAI’s free-forever plan. It includes 3 channel subscriptions.

FAQs

1. Is YouTube a reliable source for learning data science?

Yes, it can be an excellent resource, especially when you follow channels run by established professionals and educators like the ones listed. However, it’s important to practise and apply what you learn. Always supplement videos with hands-on coding and real projects to solidify your understanding.

2. I’m a complete beginner with no coding experience. Which channel should I start with?

For absolute beginners, Programming with Mosh or freeCodeCamp.org are the best places to learn Python fundamentals from scratch. A study app can help you keep a regular schedule. Once you have a basic grasp of coding, move on to Krish Naik, Alex The Analyst or Luke Barousse for data-specific projects.

3. Can these channels help me build a portfolio for job applications?

Yes. Alex The Analyst, Luke Barousse and Krish Naik are known for their end-to-end project tutorials, and Ken Jee’s older videos explain what makes a portfolio stand out. Follow along, then build your own unique project with a different dataset. Writing up what you built (repurposing it into posts for LinkedIn or a blog) also helps employers find your work.

4. Are any of these channels better for understanding the theory and math behind data science?

Yes. StatQuest with Josh Starmer is the strongest choice for an intuitive understanding of the statistics and math behind algorithms. 365 Data Science’s older videos also explain theory clearly in an academic style.

5. Which channel is best for data analytics or data engineering?

For data analytics, start with Alex The Analyst, then take Luke Barousse’s free Data Analyst Bootcamp. For data engineering, Luke Barousse’s Data Engineer Bootcamp is a good free starting point.

6. How do I decide which channel is right for me?

Consider your current skill level and learning goals. If you want to build projects, check out Alex The Analyst or Luke Barousse. If you’re stuck on a statistical concept, turn to StatQuest. If you need study or career advice, watch Tina Huang or Ken Jee’s older videos. Sample a few videos from different creators to see whose teaching style suits you best.