Statistics · Columbia University
- Theory on many statistical models such as generalized regression, different estimators, missing data, multivariate gaussians etc.
The class taught me a much more statistical perspective on machine learning
Learned important Statistical Modeling Techniques which are widely used in the FinTech industries.
I learned ML concepts more happily than in most of more maths oriented classes
Very informative, got an internship because of the concepts taught in this class !! I think there are a lot of real world applications for these concepts.
Statistical models for machine learning and real world analysis.
Maximum Likelihood, Markov Chain, Time Series, Linear Regression, General Linear Model, and so on
The class has a very steep learning curve...first several weeks I suffered a lot, especially for doing the homework problems. However, I think that is the most rewarding class I took throughout the program. Professor Li really gets where we are and tries his best to help us out. Also, Arnab is the most responsible TA I have ever seen! Don't panic if you cannot figure out the problems and concepts, he will spare no effort walking you through during his OH. The class is thereoretical, but there are many useful concepts and underlying logistics of all things happened behind the scene when you work with models and data. Definitely that's not a course that you can happily play around with wrapped code or packages and generate cool results to brag yourself, but it takes you to a higher level to understand casual inference: when you have lots of data and with no clues, where should you begin -- I think that's what we encounter when working as a data scientist more often than the former one.
A lot, but don't know if its going to be of any use
Learn many concept but never totally understanding one of them