Professor's Pet: You are just a number

PYGHACK 2018

Introduction

Professor's Pet is a numerical model that will be able to rank classes within an academic department at UIUC, using Prof. Fagen-Ulmschneider's dataset of grade point averages of offered classes.

Our algorithm is a adaption from the chess ELO rating, as we matched classes between each other as well as students against each other. We call our metric the competitiveness rating.

Our Algorithm

Input Data Models:

Purpose:

A general purpose algorithm that calculates the competitiveness index, a pseudo-elo based score that objectively determines the difficulty of a class or professor or the ability of a student to perform in their respective major.

Results

Using the average GPA of 114 past and present ECE classes, we were able to come up with a competitiveness rating for each of them, and in the graph below show that there is a trend that classes with higher average GPAs tend to have higher ratings:

Among present classes, these are some of the hardest and easiest ECE classes, with higher ratings being easier, and lower ratings being harder. With a cursory glance, we can see that some notorious classes have a low rating due to low average GPA, and advanced graduate seminars are easier:

We also took a look at the competitive rating by professor:

Appendix:

Derived update rule

Built With

• python