Hacker News Comments on
Machine Learning Foundations: A Case Study Approach
Coursera
·
University of Washington
·
3
HN comments
- This course is unranked · view top recommended courses
Hacker News Stories and Comments
All the comments and stories posted to Hacker News that reference this url.For complete newbies (but with programming experience), I would recommend this UW Coursera course to get introduced to ML Basics: https://www.coursera.org/learn/ml-foundationsEarly this year Apple acquired Turi for $200 million. It was founded by Carlos Guestrin, one of the professors who is teaching the course.
We (Class Central) are also working on a six part Wirecutter style guide to learning Data Science online. Here is part 1: https://www.class-central.com/report/best-programming-course...
Feedback would be appreciated (on the format as well as content)!
⬐ tgokhI'm a huge fan of the rest of this Coursera specialization (or was, until they started charging to submit assignments for it mid-specialization, but I digress...)Carlos and Emily do a great job diving deeper than most other online courses into the math behind different algorithms without making the math too theoretical. I'm a grad student in engineering, so I wanted to understand not only how to run these algorithms but also how they work and these courses were great for learning in a mathematically rigorous but still approachable sort of way.
The only criticism I've heard of this series is that it uses Turi/Dato/Graphlab instead of SciKit-Learn. I did the courses that exist so far using GraphLab, but I'm starting to redo the assignments using SciKit now so that I learn that toolkit as well.
⬐ dhawalhsI think they start charging after the second course.I am in the same boat as you. I am currently doing Udacity's Machine Learning Nanodegree. But I think I would have felt lost if I hadn't done the first two courses of that Coursera Specialization.
Just started, but it seems that Pandas and SciKit-Learn are very similar to Dato/Graphlab from a usage perspective.
I liked the UW Coursera class that gave a broad overview of these topics with some applications: https://www.coursera.org/learn/ml-foundationsIt's part of a Machine Learning Specialization on Coursera (5 courses + a capstone project) which goes deeper on some areas after the foundations course: https://www.coursera.org/specializations/machine-learning
I am taking this specialization and I have learned a lot so far. The material seems like it's at exactly the right level of depth (balances giving a high level overview of the field, with enough depth in specific areas to understand how things work and be able to apply them). Disclaimer: I work at Dato, and the CEO of Dato is also one of the instructors of this course.
Certainly understanding the math is very important but it is harder to get expertise on the pre-requisite math because the horizon is much bigger. I would recommend taking a case study approach and side by side learning the math stuff needed. If you are looking for an example then take a look at this, https://www.coursera.org/learn/ml-foundations/