Case study taken from Andrew Ng's Supervised Machine Learning: Regression and Classification course on Coursera.
The projects are separated into two separate parts:
-
Execute multiple linear regression and classification from scratch with only numpy to better understand the fundamental mathematical concepts.
From here, concepts underlying multiple linear regression such as feature scaling, cost function and gradient descent are discoursed in detail. -
Library of scikit-learn is used to execute similar multiple regression and classification tasks in the second part to showcase model application in practice.
"utility files" folder consists of several custom functions provided in the course to ease implementation.