PawPularity_Predictor
Devloped models for predicting visual appeal of photos of rescued animals to help the animals get adopted as soon as possible. Used various CNNs, XGBoost, Transfer Learning, K Fold, Stacking.
Created on 25th May 2022
•
PawPularity_Predictor
Devloped models for predicting visual appeal of photos of rescued animals to help the animals get adopted as soon as possible. Used various CNNs, XGBoost, Transfer Learning, K Fold, Stacking.
The problem PawPularity_Predictor solves
Millions of stray animals suffer on the streets or are euthanized in shelters every day around the world. This project is able to accurately determine a pet photo’s appeal and even suggest improvements to give these rescue animals a higher chance of loving homes. This project was made in Kaggle competition organized by PetFinder.my, a leading animal welfare platform based in Malaysia. This project can serve as a guide to shelters and rescuers around the world to improve the appeal of their pet profiles, automatically enhancing photo quality and recommending composition improvements. As a result, stray dogs and cats can find their homes much faster. This project aims to creatively improve global animal welfare with the help of Data Science.
Challenges I ran into
While building this project, the main challenges I ran into was of data cleaning, specifically the metadata of photos. Apart from that, the hyperparameter tuning of XGBoost and CNNs was quite challenging.
Technologies used