This is a flag classifier built from two seperate models. The goal is to categorize countries by the dominant religion of the population.
Licensed under GPLv3. You should have received a copy of the license with this file.
Built using a convolutional neural network Files:
- flag_labeler.ipynb
Inputs:
- Flag images
Outputs:
- Flag attributes
Built using a random forest classifier trained with k-fold cross validation Files:
- attributes.py
- classifier.py
Inputs:
- Flag atributes
Outputs:
- One of eight religion classes.
Required libraries:
- keras
- scikit