#76 overall
Statistical Learning (Hastie and Tibshirani)
Stanford
5 recommenders · 6 threads · recommended in 4 years, 2016–2023 · score 6.1
A Stanford course taught by Trevor Hastie and Robert Tibshirani based on their book An Introduction to Statistical Learning, covering core supervised and unsupervised methods such as regression, classification, resampling, tree methods and support vector machines, with R used in the labs.
Commenters call it a terrific, practical base in classical statistical learning that is worth more than many ML or deep learning courses for data science work and interviews.
Machine Learning & AIbeginnerfreevideo lectures#statistical learning#regression#classification#r#hastie#tibshirani
Recommendations by year
20062026
This is an update to a very popular text which was originally in R. Professors Hastie & Tibshirani are leading educators in statistical learning. They also have a video course following these notes in Stanford Online. Very highly recommended if learning theoretical aspects of classical ML
Hastie and Tibshirani teach a free course based on this book on Stanford's OpenEdX (https://online.stanford.edu/courses/sohs-ystatslearning-stat...). I highly recommend taking this course or reading the book before delving into ESL. IMO, ESL is excellent as a reference, but trying to learn by reading it linearly is not an optimal time investment.
Now if only a similar course existed for Wasserman's "All of Statistics..."
I disagree re: ESL and RMS, but Harrell's book is superb. It's really more aimed at biostatisticians and clinicians, though.
If the math in ESL gives you trouble, you might prefer http://www-bcf.usc.edu/~gareth/ISL/ISLR%20First%20Printing.p... (and I'm not just saying this because the first author was one of my advisers, although I do think that he and Daniela are particularly gifted teachers).
If the math in ESL is too trivial for you, there's always https://web.stanford.edu/~hastie/StatLearnSparsity_files/SLS... , which covers some graphical modeling strategies in later chapters and even kicks the tires of the autoencoder (imho perhaps the greatest recent advance in neural networks for practitioners) along the way.
Koller's course and Ng's course are also good.
Ultimately I feel like you have to get the math right or you'll never acquire the intuition that helps you design your own approaches. But you also have to put in the work.
That reminds me, tibshirani's Stanford course (accompanies ISL and ESL) is terrific. Better than those other two, actually. I wish Harrell would offer one.
You can learn R without knowing anything else. It's a good statistics package but a dated and quirky programming language.
I'd consider doing the Stanford Statistical Learning class which starts off with teaching you R. https://lagunita.stanford.edu/courses/HumanitiesSciences/Sta...
Also recommend Swirl, which is an interactive tutorial - http://swirlstats.com/students.html
Or you can go straight to Julia, which is a modern statistical package and language without R's quirks http://julialang.org/
If you're eventually going to put your project on the Web, or just want to learn programming the right way, might be better off doing it in python.
Do what feels good for you :)
Ng is a fine place to start, you get some pretty quick wins, doing MNIST from first principles within a month or two. You just need to know or get comfortable with matrix multiplication. It strikes a reasonable balance between being rigorous and approachable for a committed student at an undergrad level.
Principles of Statistical Learning is easier https://lagunita.stanford.edu/courses/HumanitiesandScience/S...
LAFF linear algebra is just starting http://www.ulaff.net/
Hinton's Neural Networks is offered in the fall https://www.coursera.org/learn/neural-networks
For my money, I wouldn't do something like Practical Machine Learning in R, because I think you'll learn more R than machine learning. I wouldn't do the Udacity TensorFlow course because I think it assumes a lot of stuff you would learn in Ng's class ... I think Ng is a fine place to start.
Based on recent (successful) job interviewing, I'd recommend people looking for a job in data science/ML to do a statistical learning course such as the Hastie and Tibshirani Stanford one [1] as a higher priority over ML/deep learning courses. It gives you a base level of knowledge in the field, and even for jobs that do deep learning, most of the technical questions will be about making sure you know the classical concepts really well.
[1] https://lagunita.stanford.edu/courses/HumanitiesSciences/Sta...
I'm a bit disappointed as I didn't see some of my favorites here :D One of them has already been mentioned, but I'll add it anyway.
- Harvard Stat 110: awesome and somewhat challenging lecture series on probability. It goes into all the probability basics, but also goes into problem solving skill very often, so the problem sets tended to be hard as I recall it. But the nice thing is that a lot of it you can find the solutions which are very well written -- and for the exams as well. Also, the lecturer Joe Blitzstein won best professor at Harvard if I'm not mistaken. https://www.youtube.com/playlist?list=PL2SOU6wwxB0uwwH80KTQ6...
- Statistical Rethinking by Richard Malkreath: man this one will make you relearn statistics. And with a heavy bayesian flavor, which if you hadn't had the chance to learn, will bend your mind as well. You will learn to build models that can describe a lot of situations in the real world, and estimate the parameters from data. Cool stuff if you ask me. https://www.youtube.com/watch?v=BYUykHScxj8&list=PLDcUM9US4X...
- Frank Harrel's Bioistatistics for biomedical reasearch: Frank Harrel is the go to guy to understand how to use data in clinical trials and diagnostics research. His book Regression Modeling strategies is a gem that every data scientist should read. This lecture series is aimed at biomedical researches, ie. people without a strong background in theoretical statistics. In the lectures he talks about the best practices and pitfalls you'll come accross when doing and reading research, and also explain some R code to do a better job. Harrel also wrote some very important R packages i.e. Hmisc and rms. https://www.youtube.com/@bbrcourse6203/videos
- calling bullshit in the era of big data: this is a last year course so it is very laid back in the discussions. I didn't go through the whole thing. But what I watched I remember it was really nice and thought provoking. https://www.youtube.com/watch?v=A2OtU5vlR0k&list=PLPnZfvKID1...
- statistical learning by Hastie and Tibshirani: these are the guys that wrote the two main books on statistical learning. If one wants to get into DS, this is the place to start. https://www.youtube.com/watch?v=LvySJGj-88U&list=PLoROMvodv4...
- Discrete Differential Geometry by Keenan Crane: ok, I didn't see the whole thing, because it was above my understanding. But the graphics and images are so eye catching I almost wanted to just sit there watching. I'm pretty sure this and his computer graphics lectures are aso engaging as hell and hidden gems of the internet. https://www.youtube.com/watch?v=mas-PUA3OvA&list=PL9_jI1bdZm...
A couple recommendations piggybacking off of yours:
A First Course in Probability has a lot of problems (with solutions) and worked examples, but it’s light on intuition and pedagogy. It’s not an easy book to learn from, on its own. I highly recommend listening to Joe Blitzstein’s STAT 110 lectures and reviewing the wealth of problems/notes. The greater mastery of probability theory that you have, the easier studying ML and stats is. https://projects.iq.harvard.edu/stat110/home
Elements of Statistical Learning is a true textbook—a comprehensive bible that could occupy you for many thousands of hours. ISLR is the better book for a crash course: http://faculty.marshall.usc.edu/gareth-james/ISL/
There are also lectures and slides from the authors: https://www.dataschool.io/15-hours-of-expert-machine-learnin...