Description

Machine Learning (UML501)
Q 1 Multiclass Logistic Regression Implement Multiclass Logistic Regression (step-bystep) on Iris dataset using one vs. rest strategy?
Q 2 Ridge Logistic Regression Download the exam dataset from the following link: https://drive.google.com/file/d/1wH6ofvNGPmORFlCLt72WGhJYPZiXstYh/view ?usp=sharing
The dataset labels that whether or not the student will get admission on the basis of the two exam scores. The plot of the data against exam1 and exam2

As clear from the figure, a linear decision boundary does not fit well. So, fit a Logistic Regression Classifier with polynomial function of test1 and test2 scores upto degree
6 using
i. Step-by-Step Logistic Regression (with no regularization; alpha=10; number of iterations=1000)
ii. Step-by-Step Logistic Regression (with ridge regularization; alpha=10; number of iterations=1000; lambda=0.2)