9 Best YouTube Channels for Data Science in 2026 (Ranked)

Quick Answer: The Best Data Science YouTube Channels in 2026

The best YouTube channels for learning data science in 2026 are StatQuest for statistics, Alex The Analyst for the data analyst toolkit, Luke Barousse for practical analytics, 3Blue1Brown for the underlying mathematics, and Krish Naik for end-to-end machine learning projects. sentdex, Ken Jee, Data School and Rob Mulla round out the list. Every channel below was checked live in September 2026: the subscriber figures and the last-upload dates are readings taken this month, not carried over from an older version of this page. Three of the nine have gone quiet, and this page now says which ones and what they are still good for.

The full ranked list is below.

Channel Comparison Table

Ranked by a combination of teaching quality and verified publishing activity. Every subscriber figure and last-upload date in this table was read from the live channel page in September 2026.

The Data Science YouTube Problem (And How These Channels Solve It)

Most data science YouTube content falls into one of two traps, and the channels below were selected because they avoid both.

The first trap: it is too easy. A clean dataset loads perfectly, a model trains in seconds, accuracy is 96%, everyone claps. You watch it, feel great, try to apply it to real data, and spend four hours debugging a CSV encoding error.

The second trap: it is too narrow. Machine learning theory in isolation, or SQL in isolation, or statistics in isolation. Data science requires all of these working together, and channels that cover only one piece leave you with a partial picture.

The selection criteria for this list were:

• Practical accuracy - does the content reflect how data science is done at companies, not just in textbooks?
• Skill coverage - does it address the full stack: statistics, Python, SQL, visualization, ML, communication?
• Career relevance - does it help you get and do a real data science job?
• Teaching quality - after watching, do you understand something, or just feel like you have watched something?
• Verified currency - is the channel still publishing, and if not, is its material the kind that stays true?

That last criterion is why this page now carries a last-upload column. Three of the nine channels here have not posted in five months or more.

Currency Check: Which of These Channels Are Still Publishing

Four of the nine channels on this list published within the week this page was updated: StatQuest, Alex The Analyst, Krish Naik and sentdex. 3Blue1Brown is about two months out and Luke Barousse about three, both normal for creators who publish long-form work. Three have gone quiet for much longer. Ken Jee last posted about eight months ago, Rob Mulla about six months ago, and Data School about ten months ago. None of the three has been cut, because statistics, pandas, scikit-learn and portfolio judgement do not expire the way tooling does. But you should know which is which before you plan a study schedule around them.

The practical rule: use the archived channels for fundamentals and the active ones for anything touching current tooling, hiring or AI workflows. Ken Jee's portfolio playlists are still the clearest statement of what a hiring manager looks for, but his read on the 2026 job market is a year out of date. Data School's pandas and scikit-learn series teach APIs that have barely changed. Rob Mulla's feature engineering material is still the best free Kaggle content on YouTube even though the competitions he references have closed.

Two of the active channels have also drifted. sentdex and Krish Naik both publish frequently, but their recent uploads are largely about large language models, agents and local AI rather than the classical data science their back catalogues cover. That back catalogue is still there and still good. Just do not expect the front page of either channel to look like a data science curriculum.

The 9 Best Data Science YouTube Channels

1. StatQuest with Josh Starmer - Best for Statistical Foundations

Subscribers: ~1.7M | Last upload: same week as this update | Focus: Statistics, probability, machine learning algorithms

StatQuest is the best statistics and machine learning education channel on YouTube, and it is also the most active channel on this list. Josh Starmer's method is to take a concept that intimidates most people, such as PCA, logistic regression, gradient boosting or regularization, and break it into the smallest possible pieces until each piece is obvious. Then he reassembles them.

What makes StatQuest different is that Josh never shortcuts the math. He shows you why formulas work, not just that they work. This matters in data science, where you will regularly hit a model behaving strangely and need the underlying statistics to diagnose it. Practitioners who skipped the stats and went straight to scikit-learn function calls get stuck on exactly these problems.

His catalogue covers classical ML thoroughly: linear and logistic regression, decision trees, random forests, SVMs, cross-validation, regularization and more, organised into topic playlists. He has continued adding neural network and transformer content using the same methodical approach, and was still publishing new explainer videos the week this page was updated.

Best for: Anyone who wants to understand statistics and machine learning algorithms properly. Required viewing for data scientists at every experience level.

Start with: A Gentle Introduction to Machine Learning to see his teaching style, then work through the Statistics Fundamentals playlist before the Machine Learning one. The Neural Networks / Deep Learning playlist comes after those.

2. Alex The Analyst - Best for the Full Data Analyst Toolkit

Subscribers: ~1.4M | Last upload: within a week of this update | Focus: SQL, Excel, Tableau, Power BI, Python, job prep

Alex Freberg's Data Analyst Bootcamp, published free on his channel, is the most useful structured free resource for getting started in data analysis. It sequences the tools that appear most consistently in data analyst job descriptions, with enough depth to be job-ready in each.

What Alex does well is teaching tools in context. His SQL material does not just explain syntax, it covers the queries analysts actually run: window functions, CTEs, joins across multiple tables and aggregate queries for summary reports. The visualization tutorials show how to build dashboards that communicate something rather than dashboards that merely display data.

His career content is practical too. He has documented his own move from a non-technical background into data analysis, which makes the advice more credible than someone who went straight from a computer science degree into the field. Note that the channel now mixes tech-news commentary in with the tutorials, so the front page is less curriculum-like than it used to be. The playlists are where the courses live.

Best for: Beginners to data analysis who want a structured path through the core toolkit, and career-changers who need a job-ready skillset quickly.

Start with: the Data Analyst Bootcamp from the beginning. If you only need the query skills, SQL Basics for Data Analysts stands alone, and the Tableau and Power BI tutorial series cover the visualization half.

3. Luke Barousse - Best for Data Analytics in Practice

Subscribers: ~670K | Last upload: about 3 months before this update | Focus: Data analytics, SQL, Python, job-posting analysis

Luke Barousse comes from a data analyst background and his content reflects what data science looks like in most companies: a lot of SQL, a lot of Python for data manipulation, some visualization, and constant questions about what the numbers mean. His recommendations are grounded in job posting data he analyses himself rather than in general advice.

His Python for Data Analytics full course is one of the most practical free data science Python resources available. It is not abstract: it uses real datasets, teaches pandas and visualization in context, and keeps connecting back to why you would do this at work. The SQL course is similarly practical, covering the queries you will actually write rather than academic edge cases.

He has also built out full free courses in Excel and Power BI and a multi-hour Data Analyst Bootcamp, so a learner who wants one creator's voice across the whole analyst toolkit can get it here. His most recent uploads have extended into data engineering territory with a free bootcamp on that too.

Best for: People pursuing data analyst or data science roles who want to know which skills to prioritize based on real job market data.

Start with: Python for Data Analytics, or SQL for Data Analytics if you need query fundamentals first. The Data Analyst Bootcamp: Zero to Hero playlist bundles the full sequence.

4. 3Blue1Brown - Best for Mathematical Intuition Behind Data Science

Subscribers: ~8.6M | Last upload: about 2 months before this update | Focus: Linear algebra, calculus, neural networks, visual math

Grant Sanderson's 3Blue1Brown is not a data science channel. It is a mathematics education channel, and it earns its place here because the mathematics underlying data science is where most practitioners have the biggest gaps.

Linear algebra is everywhere in data science: PCA, linear regression, neural network computations, dimensionality reduction. Most people who use these techniques treat them as black boxes. The Essence of linear algebra series makes the underlying geometry intuitive in a way textbooks rarely achieve, and the Neural networks series does the same for backpropagation and gradient descent.

Understanding these foundations changes how you think about models. Instead of asking which algorithm to try next, you start asking what your data looks like in high-dimensional space and which technique is designed for that geometry. That is a qualitatively different kind of reasoning, and it is what separates someone who can run scikit-learn from someone who can debug it.

Best for: Data scientists at any level who want to actually understand the math they are applying rather than just use it.

Start with: Essence of linear algebra in full, then the Neural networks series from chapter one.

5. Krish Naik - Best for End-to-End Machine Learning Projects

Subscribers: ~1.5M | Last upload: same week as this update | Focus: ML projects, MLOps, deployment, generative AI

Krish Naik covers machine learning at every stage from raw data to production deployment, and his channel is one of the most complete on YouTube in terms of scope. Statistics, Python, SQL, machine learning, deep learning, NLP, model deployment and MLOps are all covered with substantial depth.

The end-to-end project videos are the reason to come here. Most tutorials stop at model training. Krish's full walkthroughs continue through preprocessing pipelines, model serialization, API creation, containerization and deployment. Seeing the complete arc of a real project in one place is worth more than five separate tutorials on individual pieces.

One caveat worth stating plainly: the channel's recent output is dominated by generative AI, agent frameworks and retrieval pipelines rather than classical data science. That is where his new playlists are going. The classical machine learning and MLOps material is still published and still good, but you will need to navigate to it through the playlists rather than the home page.

Best for: Intermediate learners who want to understand the full lifecycle of a machine learning product, from exploration through to deployment.

Start with: the End To End Data Science Playlist for industry-shaped projects, then MLOPS for the deployment half. The Machine Learning 2024 playlist is the classical-algorithms entry point. A handful of videos inside his project playlists are behind YouTube channel membership; the bulk of both playlists is free.

6. sentdex - Best for Python Data Science with Real Data

Subscribers: ~1.4M | Last upload: same week as this update | Focus: Python, pandas, deep learning, applied AI

Harrison Kinsley's sentdex teaches programming the way it actually happens: messy data, unclear objectives, tools that do not quite cooperate, and solutions assembled from libraries that were not designed for each other. Where other channels show the cleaned, idealized version, sentdex shows the real one.

His back catalogue covers the Python data stack properly. There are full series on pandas-based data analysis, on deep learning with TensorFlow and Keras, on building neural networks from scratch, and on learning Python itself from the beginning. The emphasis on doing rather than explaining means the pace is faster than other channels, but you see complete working pipelines on actual data.

Be aware of the drift, though. The channel's current uploads are almost entirely about local large language models and AI harnesses rather than data science. If you come to sentdex for pandas and scikit-learn you will find excellent material, but you will find it in the playlists rather than in anything published recently.

Best for: Python programmers who want to apply their skills to data problems and see how things work on real, imperfect data.

Start with: Data Analysis w/ Python 3 and Pandas. If Python itself is new, Learning to program with Python 3 comes first, and Deep Learning basics with Python, TensorFlow and Keras or Neural Networks from Scratch after.

7. Ken Jee - Best for Data Science Career Reality (Archive)

Subscribers: ~280K | Last upload: about 8 months before this update | Focus: Career path, portfolio projects, job search

Ken Jee is the most honest voice on YouTube about what a data science career actually looks like. Where most channels cover only the technical side, Ken covers what determines whether you get hired: building a portfolio that stands out, preparing for interviews that are nothing like LeetCode grinding, communicating results to non-technical stakeholders, and understanding what separates a junior from a senior.

His portfolio review playlists are the most useful thing here. He critiques real subscriber portfolios and shows what a hiring manager sees when they open your GitHub. The gap between "I think this is impressive" and "this is what actually impresses hiring managers" is usually wide, and watching him close it on someone else's project is more instructive than any checklist. He also started the #66DaysOfData challenge, a public commitment to work on data science skills every day for 66 days, though the playlist he collected it under is now down to a single clip, so look for the individual videos rather than the playlist.

The channel has not published in roughly eight months, and his final run of uploads was about AI disruption and product building rather than data science careers. Treat it as an archive: the portfolio and interview material holds up, the read on the current job market does not.

Best for: Aspiring data scientists who want to understand what the job involves and what a portfolio needs to show. Use it for judgement, not for current market conditions.

Start with: Build a Data Science Portfolio, then Reviewing Your Data Science Projects to see the same standards applied to real work.

8. Data School (Kevin Markham) - Best for pandas and scikit-learn Deep Dives (Archive)

Subscribers: ~260K | Last upload: about 10 months before this update | Focus: pandas, scikit-learn, Python workflow

Kevin Markham's Data School covers pandas and scikit-learn in more depth than any other channel on this list. His pandas best-practices material covers the API practitioners actually use, the mistakes beginners make, and the patterns that hold up on larger datasets. If you learned pandas from a quick tutorial and know something is missing, this is what fills it in.

The scikit-learn tutorials are similarly deep, covering preprocessing pipelines, cross-validation strategies, model selection and reproducible workflows. The emphasis on correct validation is important and routinely skipped in faster-paced tutorials. He also worked through the ISLR statistical learning textbook chapter by chapter, which is a rare free companion to a standard text.

The channel stopped publishing about ten months ago and its last uploads were AI news rather than tutorials. This matters less here than almost anywhere else on this list, because the pandas and scikit-learn APIs he teaches have been stable for years. Use it as a reference, not as a source of current practice.

Best for: Data scientists who already have basic Python and want to use pandas and scikit-learn the right way rather than just the working way.

Start with: Best practices with pandas if you already know the basics, or Data analysis in Python with pandas if you do not. Then Machine learning in Python with scikit-learn.

9. Rob Mulla - Best for Competitive Data Science on Kaggle (Archive)

Subscribers: ~220K | Last upload: about 6 months before this update | Focus: Kaggle, EDA, feature engineering

Rob Mulla runs the best Kaggle-focused data science content on YouTube. It covers the full competition workflow: exploratory data analysis, feature engineering, model selection, hyperparameter tuning and ensembling. The approach is unfiltered, showing his actual notebooks including the dead ends.

Kaggle competitions are one of the better learning environments for data scientists because they give you standardized problems, real datasets and immediate feedback. Rob's material teaches you not just how to enter but how to improve, and the difference between a top-50% and a top-10% submission is usually feature engineering rather than algorithm choice. His live-coding and EDA walkthroughs, where he explores an unfamiliar dataset and narrates his thinking, teach the judgement calls that separate good data scientists: what to look for, what to worry about, which patterns matter and which are noise.

Uploads have slowed to roughly one every several months, with the most recent about six months before this update. The workflow material does not depend on that, but the specific competitions he works through have long since closed.

Best for: Data scientists who want to level up through competition work and learn professional-grade EDA and feature engineering.

Start with: A Gentle Introduction to Pandas Data Analysis (on Kaggle) to see his process, then the Working with Data in Python and Machine Learning Tutorials playlists.

How to Structure Your Data Science Learning

Phase 1: Statistics and Python Foundations (Weeks 1-8)

Start with StatQuest's Statistics Fundamentals playlist, working through probability and classical statistics before touching machine learning. In parallel, learn Python with pandas through Luke Barousse or sentdex, and add SQL through Alex The Analyst. These three form the actual core of most data science work: not deep learning, not fancy models, but the ability to query data, manipulate it in Python, and know which statistical test applies.

Phase 2: Machine Learning Fundamentals (Weeks 9-16)

Continue with StatQuest's Machine Learning playlist and learn the algorithms properly, including what they assume and where they break. Implement them in scikit-learn using Data School's workflow tutorials. Supplement with 3Blue1Brown's linear algebra series for the mathematical foundation. Complete two or three Kaggle getting-started competitions using Rob Mulla's EDA approach.

Phase 3: Full-Stack Data Science (Weeks 17-24)

Add deep learning through StatQuest's neural networks playlist or sentdex's TensorFlow series. Extend your SQL and visualization through Luke Barousse's and Alex The Analyst's Power BI and Tableau courses. Build projects that go end to end using Krish Naik's project playlists. Focus on portfolio quality following Ken Jee's review standards. By the end of this phase you should have three or four portfolio projects that demonstrate the full lifecycle.

Phase 4: Specialization and Competition (Ongoing)

Pick a specialization: NLP, computer vision, time series, or a specific industry domain. Use Rob Mulla's feature engineering material to compete on Kaggle in that area. For anything touching current AI tooling, the active channels are the ones to follow, since Krish Naik and sentdex both publish regularly in that space.

How LearnPath Builds Your Data Science Path

Data science has a particularly severe curriculum problem. The field is broad, the resources are scattered across statistics, Python, SQL and machine learning channels, and the correct sequence is unclear for most learners. You can spend months on ML algorithms before you have the Python or statistics foundation to apply them correctly.

Our own numbers show how that plays out. Across the paths our learners have built on data science, data analysis, statistics, pandas or scikit-learn topics, we count 30 paths from 25 learners with 117 videos queued, an average of 3.9 videos per path. Only 6 of those videos have been watched past the one-minute mark and only 2 have been finished, across a single path. That is a small sample and we are not going to dress it up as more than it is, but it matches the pattern we see on every broad topic: queueing a curriculum is easy and finishing the second video is not.

LearnPath is built around that gap. Tell it your current skills and your goal, and it builds a learning tree from YouTube content, sequences it, generates quizzes from the transcript of the video you just watched to check that you actually understood it, and branches based on the result.

Skip the curation. LearnPath builds your path automatically.

Frequently Asked Questions

How long does it take to become a data scientist from scratch?

With consistent daily effort, 12-18 months to be competitive for entry-level positions. Data analyst roles are achievable in 6-9 months with strong SQL, Python and visualization skills. The timelines are highly variable based on math background, programming experience, and how much time you can dedicate each week.

Do I need a degree for data science?

Not necessarily, but a degree in statistics, mathematics, computer science or a quantitative field is a real advantage, especially at larger companies. For smaller companies and data analyst roles, a strong portfolio and demonstrable skills often matter more. The fastest path without a degree is a focused portfolio of real projects and strong SQL and Python skills.

Should I learn Python or R for data science?

Python for most industry roles in 2026. R stays strong in academic research and specialized domains like clinical trials and econometrics, and both Alex The Analyst and Data School keep free R material online. Python's data ecosystem, pandas, scikit-learn and PyTorch, is the one most of these channels teach against, so start there and add R only if a specific role requires it.

Does it matter if a data science channel has stopped uploading?

It depends what you want from it. Statistics, linear algebra, pandas and scikit-learn fundamentals do not expire, so an archived channel is still a good teacher for those. Tooling, hiring advice and anything touching current AI workflows go stale fast. Check the last-upload column above before you plan a schedule around a channel.

What SQL should a data scientist know?

Joins across multiple tables, aggregate functions and GROUP BY, window functions (ROW_NUMBER, LAG/LEAD, rolling aggregates), CTEs for query organization, and subqueries. These cover about 90% of what you will write as a data analyst or scientist. Alex The Analyst and Luke Barousse both teach these patterns in free full-length courses.

How do I build a data science portfolio without work experience?

Kaggle competitions, personal projects on topics you find genuinely interesting, and contributions to open-source data projects. The key is choosing projects that involve messy real data and require the full workflow: collection, cleaning, exploration, modeling and a clear conclusion. Ken Jee's portfolio playlists show what the bar is.

Where to Start

The shortest path to data science skills:

1. StatQuest's Statistics Fundamentals for the statistics you cannot skip
2. Luke Barousse's Python for Data Analytics course
3. Alex The Analyst's SQL Basics for Data Analysts
4. Ken Jee's portfolio playlists for what the finished work needs to look like

If you want those resources sequenced into a structured, adaptive learning path with real comprehension checks, LearnPath builds it automatically.