Assignment 0

Due in class at 1:30pm, July 9. You may work in a group of up to three students, but each individual \emph{must} be involved in every question. Do not assign problems to individuals within a team! Please submit only one solution per group.

The PGSS admissions committee has decided to consider applying machine learning techniques to accepting students. They gathered records of previously admitted students, giving students scores based on their SAT, GPA, and essay. They also gave each student a label of 1 if the student performed well at PGSS and -1 otherwise. They want you to try different techniques and tell them whether they should accept Krunch. Here's the data they gathered.

1. Try linear regression on this data, inserting the additional constant-1 attribute as in the text.

   Show \emph{all} your work in this and all the following problems. In this problem, be sure to clearly indicate the set of equations you find to solve, the hypothesis function you discover, and the resulting verdict for Krunch.

2. Try nearest-neighbor search using the Euclidean distance function. Clearly indicate a table showing the distance to each student and the resulting verdict for Krunch.