Austin_AniML_Rescue was a machine learning project that took data from the Austin Animal Center in order to determine what factors most impacted the outcomes of cats and dogs who passed through the shelter. Our project started out open-ended, where we could choose and dataset we wanted in order to analyze any problem that we wanted.
This page was written to give further insight into my personal contributions to the Austin_AniML_Rescue project, a machine learning-focused project that I collaborated on with my teammates Ilaha Cardoso, Sarah Griffin, Melinda Malone, and Ruth Ordeo.
To see the full, original project, please visit the link here: https://github.com/ilaha/Austin_AniML_Rescue
The files included in this repository are the ones that I worked on. With the exception of the merged_csv_cleaning.ipynb file which had some contributions from Ilaha, I created and scripted 100% of everything else seen here. I had the critical role of developing the machine learning models, as well as performing the initial exploratory data analysis, and cleaning/prepping the data to feed into the models.
Relevant Files
- Austin_AniML_EDA.ipynb
- intakes_outcomes_EDA.xlsx
I performed the initial exploratory data analysis for the team in order to understand how the data was organized and what the main categories of interest were. I used jupyter notebook and pandas to see a break down of the statistics, and then created an Excel workbook to create visualizations of my findings that I shared with the rest of the team.
From the analysis, I learned that the vast majority of animals in the shelter were cats and dogs, and that the main outcomes of significance for them were adoption, transfer, return to owner, and euthanasia. Based on this information, our team decided to narrow our focus down to just cats and dogs, and to examine the outcomes -- adoption, transfer, and euthanasia.
Relevant Files
- merged_csv_cleaning.ipynb
Relevant Files
- Logistic_Features_Cats_Dogs.ipynb
- Machine_Learning.ipynb
- Sampling_Euthanasia.ipynb
- Top_Features_Cats_Dogs.ipynb
- XGBoost_Features_Cats_Dogs.ipynb
I tested a variety of machine learning algorithms by feeding them different combinations of data.