Our goal for this final project was to create a machine learning model to improve pet adoption rates. We accomplished this through image classification techniques to analyze and recommend existing images to understand what features of an image yielded a successul adoption. Our final report can be found here.
This was part of a competition on Kaggle.com: Experiments and code for the Petfinder Pawpularity competition.
These are a lot of files in this repository. We ran a ton of experiments and most of them didn't prove helpful. Instead of reading through every journal, here is a mapping from the experiments discussed in the report to the Jupyter Notebooks containing the code.
| Experiment | Author | Notebook |
|---|---|---|
| EDA & Feature Selection | Autumn Rains | link |
| PCA | Autumn Rains | link |
| Color Palette | Carlos Moreno | link |
| Bucketing | Carlos Moreno | link |
| Image Classification | Chandler Haukap | link |
| Species Classification | Chandler Haukap | Dimensionality Reduction, Analysis |
| Eigenfaces | Chandler Haukap | Scaling and Padding, PCA, Analyis |
| Feature Selection | Carlos Moreno | link |
| Model Creation, Evaluation and Selection | Carlos Moreno | link |