CERBERUS - IEEE HOST 2023

DEMO LINK

This repository replicates (including images) the environment we used to train and run our model for the 2023 IEEE HOST SCS competition.

There are two important python notebooks:

• build_model.ipynb
• run_model.ipynb - USE THIS TO RUN OUR MODEL

build_model.ipynb

This file contains the code we used to train and test our original model.

run_model.ipynb

Merely open the notebook and run.

This notebook will run the model on:

• the training data
• the test data
• the holdout data

For the test and training data, the accuracy and confusion matrix will be printed.

For the holdout data, a list of predictions and their associated IDs will be printed. This should match the outputs shown in sample_submission.csv.

Submission Information

• The report is provided in this folder at report.pdf
• Code is provided at https://github.com/amacgillivray/cerberus_HOST_2023_SCS
• Code without data is also available in "code_nodata" in this directory
• Sample submission is in "sample_submission.csv"
• The demo is provided at https://youtu.be/JxXAU5iE-uA
• The presentation is provided in this folder, as both a powerpoint and PDF:
  • presentation.pptx
  • presentation.pdf