Project 2

Contents

Overview

Project 2 is a continuation of Project 1. You will use the same dataset you used in Project 1 and identify an interesting predictive task for the dataset that fits in nicely with the story you developed in Project 1. The goal of Project 2 is for you to demonstrate your ability to identify predictive tasks, choose two relevant predictive models for that task, evaluate them, and analyze the results. I will provide you with a held out portion of your dataset that will be used to test your models after you've tuned them.

You will present your results of Projects 1 and 2 during the finals period for this class in May. Presentations should be done in PowerPoint (or similar) and should contain graphics from your Jupyter notebook.

Datasets

As described above, the same datasets will be used as in Project 1. However, I will provide you with an additional, held out portion of your dataset after you have tuned your models (send me an email once you have done that).

Project constraints

Your project will be evaluated on the following criteria:

• You've folded in any comments I made on your Project 1
• You've identified an interesting prediction task that fits in with the story you developed in Project 1
• You explored at least two models that were appropriate for the task you identified
• You tuned and tested the models appropriately and analyzed the results
• All processing steps are contained within the submitted Jupyter notebook and are well documented; I must be able to reproduce your steps
• Your oral presentation of your story includes a slide show of some sort, is easy to follow, and is engaging

See the rubric on the submission page for a break down of how these constraints will be graded.

Submissions

See this Canvas page to submit the project code and see the rubric.

See this Canvas page to submit the presentation code and see oral presentation rubric.