Re:Machine Learning Crash Course: Part 1
by sickrumbear on Thursday December 29, 2016 @02:44PM
(#13279504)
Are there any resources you would suggest for someone in the same boat to get easily reacquainted and rebuild a good foundation?
Re:Machine Learning Crash Course: Part 1
by JshWright on Thursday December 29, 2016 @02:59PM
(#13279642)
The Coursera Machine Learning course just started (I assume you could still join). I just finished the second week (I'm trying to keep a week ahead due to the somewhat unpredictable nature of my schedule lately), and have been enjoying it so far.
The first couple weeks are all about univariate and multivariate linear regression (as well as an optional linear algebra refresher on matrix operations).
Re:Machine Learning Crash Course: Part 1
by cr0sh on Thursday December 29, 2016 @04:27PM
(#13280269)
I second this course; I took it when it was put on in 2011 in association with Stanford, and called the "ML Class" - its success was the catalyst for the creation of Coursera.
Re:Machine Learning Crash Course: Part 1
by abhgh on Thursday December 29, 2016 @11:53PM
(#13282607)
I occasionally organize ML "orientation" sessions, and one of the first models I talk about is linear regression. For this very reason. Its an excellent bridge between what people are already familiar with and the larger "learn from data" philosophy.
As a data scientist, I often walk new clients through a linear regression exercise to convey some key concepts about the engagement and demystify what I'll be doing for them.
I'm often dealing with people who, much like you, haven't done much with stats since a college "Business Stats" course, so I get a lot of "oh yeah, I vaguely remember this" - but going through it again gives me a good foundation to relate back to as things get more complicated.