REG - 3 | Better computation methods for regression

After a cursory look at everyone’s projects, I see that we’ve implemented inversion methods to invert X and to find optimal solutions to our linear regression problems. This, however, is not computationally optimal. There exist methods like QR decomposition, Cholesky, and SVD Decomposition, that make it more computationally efficient and stable to solve the linear regression.

Let’s dedicate this week to studying such methods and their merits:

1 Like

https://www.youtube.com/watch?v=WWe8yVccZy0&list=PLxKgD50sMRvBHxvNPnGQ1kEHlO5y7mSnh&index=6 found this series to be especially useful for understanding QR decomposition.

1 Like