This repo contains Julia Rodd's assignments from Northwestern University's MSDS Artificial Intelligence class.
This class featured four assignments. Below is a brief summary of each assignment:
- Assignment 1 - Leverages the Scikit Learn digits data set to understand how basic neural networks (MLP) work.
- Assignment 2 - Explores various CNN model architectures using black and white images of dogs and cats.
- Assignment 3 - Explores various RNN model architectures using a reduced reuters corpus with one topic per document.
- Assignment 4 - ompares/contrasts various model architectures using longitudinal data from a call center.
All assignments leverage the package keras with tensorflow backend.
Another uncommon package that might need to be installed is PrettyTable to render the tables used in these assignments.
All code files can be found in the 'code' folder. Jupyter notebooks were used for all assignments.
Relevant data can be found in the 'data' folder.
Assignment summaries can be found in the 'output/reports' folder.
There are some limitations in storing the data that is used in some assignments.
- Data for assignment 2 is not provided.
- Data for assignment 4 has to be downloaded first from Kaggle.
- Julia Rodd