Zoo Animal Classification

Do you remember classifying animals in the childhood? We do it all the time, probably in kindergarten, in a field trip, or in a biology class. Animal classification is so important to help us understanding the nature. If we could build a web app that predicts what class an animal belongs to, it could be beneficial in educational settings such as in a biology class or in a zoo. Furthermore, when we encounter a new animal, we could use this app as an auxiliary to predict its species. In this project, we used a dataset that contains information about 214 zoo animals with their characteristics and classes. We employed scikit-learn to train a Random Forest Classifier model, and used it to predict the class of any animal given its characteristics. To maximize the accuracy, we tuned our hyperparameters utilizing randomized search and grid search. We achieved a final accuracy of 92.1875%, with low variance. Try our webapp in the GitHub link below!

Built With

• flask
• python
• scikit-learn