Mini-Project

To help everyone understand the value of these tools, we’ll spend two class days working on “mini-projects”. Students will form groups of two or three to:

• obtain a (clean) dataset

  • Kaggle

• apply a tool in our curriculum:

  • Perceptron

  • Line of Best Fit

  • Dynamical System

  • Hypothesis Testing

  • Regression (Statistical Significance)

  • Bayesian Network

• write a brief discussion which provides

  • an examination of whether a method’s assumptions are satisfied in this application

  • discusses whether analysis outputs are trustworthy

    • (its very possible to do correct math and produce nonsense results: maybe the dataset is biased?)

  • offers a quick summary of results which is easily understood by non-technical readers

Note

You needn’t have any prior Python programming experience to succesfully complete these tasks.

The mini-projects will be graded on the mathematical understanding of the method used, not their implementation. Additionally, students will have access to:

• TA support in class, and all their Python expertise

• example projects of each type to work from

It is expected that each student (of each group) gets a working Python and Jupyter Notebook installation up and running:

• Python & Jupyter Install

We hope that this experience helps students to draw the connection between the abstract “whiteboard math” and how these tools are really used.