Which are some good online platforms to learn more about Math?
April 20, 2019 1:45 PM Subscribe
I'm getting into things like machine learning and HPC. These fields, naturally require a lot of math, mainly stuff like linear algebra. I took classes on linear algebra, calculus, probability and statistics. I don't really remember much of it and while I have an inkling of what is going on, my understanding of almost everything except for linear algebra is not very good.
I've been looking at how to understand math better. The way I learn best is through videos and having a real person explain it to me. Books are fine and everything but It's really hard for me to follow a book or not fall asleep reading it. I tried Khan Academy, but it's not just doing it for me and I wasn't very fond of how they were going through the lessons, I thought it was a bit disorganized and that sometimes it started going off topic. I also thought they had far too few exercises.
I've seen that Udemy offers courses on calculus, linear algebra, probability and statistics, but I don't know that it is the best place to learn that. Coursera's quiz, assignment model is very good, but I'm not sure that I found the right courses there. If anyone can help me figure out what is best I would appreciate it. Paid or unpaid, it's all the same to me.
I've been looking at how to understand math better. The way I learn best is through videos and having a real person explain it to me. Books are fine and everything but It's really hard for me to follow a book or not fall asleep reading it. I tried Khan Academy, but it's not just doing it for me and I wasn't very fond of how they were going through the lessons, I thought it was a bit disorganized and that sometimes it started going off topic. I also thought they had far too few exercises.
I've seen that Udemy offers courses on calculus, linear algebra, probability and statistics, but I don't know that it is the best place to learn that. Coursera's quiz, assignment model is very good, but I'm not sure that I found the right courses there. If anyone can help me figure out what is best I would appreciate it. Paid or unpaid, it's all the same to me.
I really liked Introduction to Statistical Learning for understanding machine learning algorithms and statistics. This book is also extremely clear, and extremely realistic, about the statistics used in data science. Both books also couple their lessons with R examples, which is handy.
posted by thebots at 2:32 PM on April 20, 2019
posted by thebots at 2:32 PM on April 20, 2019
Oh, I just re-read your question and saw you prefer videos. Andrew Ng's course on Machine Learning is a popular one and it definitely helped me.
posted by thebots at 2:36 PM on April 20, 2019
posted by thebots at 2:36 PM on April 20, 2019
Response by poster: The guy who does 3blue1brown was a Khan Academy content creator before he struck out on his own to try a different style. His "just as himself" content on that channel uses a lot of Python-driven animations plus narration . The source code for the tool he wrote to generate the videos is on github too, so you could follow along with not just the math ideas, but also the software approaches. He's really good at explaining complex mathematical concepts in an approachable way
I have seen some of his videos but not in any particular order, they were helpful. I will take a look.
Oh, I just re-read your question and saw you prefer videos. Andrew Ng's course on Machine Learning is a popular one and it definitely helped me.
I think it's the best introductory course. The others either simply go in too deep, or do not explain enough. Andrew Ng's course has a balance of both.
posted by Braxis at 2:52 PM on April 20, 2019 [1 favorite]
I have seen some of his videos but not in any particular order, they were helpful. I will take a look.
Oh, I just re-read your question and saw you prefer videos. Andrew Ng's course on Machine Learning is a popular one and it definitely helped me.
I think it's the best introductory course. The others either simply go in too deep, or do not explain enough. Andrew Ng's course has a balance of both.
posted by Braxis at 2:52 PM on April 20, 2019 [1 favorite]
MIT OpenCourseWare is videos of class lectures. https://ocw.mit.edu/courses/audio-video-courses/
It's free so easy to try. That said, the usual way to learn math is to do the exercises at the end of the chapter. That presupposes finding the right book, of course.
posted by SemiSalt at 4:11 PM on April 20, 2019 [2 favorites]
It's free so easy to try. That said, the usual way to learn math is to do the exercises at the end of the chapter. That presupposes finding the right book, of course.
posted by SemiSalt at 4:11 PM on April 20, 2019 [2 favorites]
Best answer: Yes, if you're specifically looking for targeted videos:
posted by gwint at 8:01 PM on April 21, 2019
- As mentioned above, 3Blue1Brown Essence of linear algebra is very good. It's taught mainly using a visual geometric approach. He also has a brief course specifically on Neural Networks
- Better Explained is another good source. He also has blog posts which are very visual, like A Programmer’s Intuition for Matrix Multiplication which is useful for ML
- You may also like the Computerphile series on neural networks and their other videos
- Mathematics for Machine Learning is more like a Khan Academy playlist but more focused on getting you to where you need to be for ML rather than SAT prep like Khan Academy
posted by gwint at 8:01 PM on April 21, 2019
This thread is closed to new comments.
posted by Alterscape at 2:02 PM on April 20, 2019 [4 favorites]