This repo contains the code and supporting documents I created for my final project in a graduate class on advanced machine learning. This program tested the effectiveness of transfer learning methods to improve the performance of a plant disease classification model. For more information on my motivation for this project and the details of its construction, read the project report included in this repo.
- "Final Report" PDF - Comprehensive report on the project and its results (contains the most information on the project).
- "Final Presentation" PDF - Final presentation that walks through report highlights with graphics.
- data_processing.py - used to pre-process the data used in the experiment.
- driver.py - main script used as a platform for hyperparameter tuning.
- fromIN.py - used to train models from ImageNet weights.
- fromscratch.py - used to train models from random weights.
- model.py - used to set up testing/validation sequences and compile results.