Tomorrow I start a Machine Learning Class at Columbia University through their executive eductation program. The program is a partnership between Columbia and a company called Emeritus. There are lots of great, free online resources so you might ask why I chose to pay for this course. In my experience, the classes I’ve seen online have either been too rudimentary for me or way beyond what I can do. There’s a gap for people like me who’ve been practicing engineers for 2 decades who have a strong mathematical background. This class looks like it fills that gap for me.
Overall the class is about $2469 so it’s not cheap. Emeritus does allow you to make monthly payments for an extra $100 so I chose this option.
MIT has a program called Pro-X which is about the same price but it’s only 6 weeks long and not as comprehensive. The MIT program looks good but I didn’t choose it because I thought I would get more bang for my buck from the Columbia course.
You can see the course overview here:
The curriculum was developed by John Paisley and while his appointment is in EE and not in CS, his CV shows he’s quite accomplished in the areas of machine learning for image processing. I’ll be posting here as to what the teaching is like. I suspect that I’ll not actually have any interaction with John and there will be course instructors who will do all the heavy lifting.
The class is 24 weeks and is divided into 2 sections:
- Python for Data Science in Weeks 1 through 12
- Applied Machine Learning in Weeks 13 through 24
- Probability and Statistics (4th Edition) by Morris H. DeGroot and Mark J. Schervish
- Python from a Data Science Perspective
- Classic Texts
- T. Hastie, R. Tibshirani and J. Friedman, The Elements of Statistical Learning, Second Edition, Springer
- C. Bishop, Pattern Recognition and Machine Learning, Springer