/Gender-Prediction

Predicting Gender using preference and catboost model

Primary LanguageJupyter NotebookMIT LicenseMIT

Gender prediction App

This is web app powered by flask and Heroku that predicts gender based on preference. The model that had the higher accuracy score was the Catboost model which was saved and used underneath the app to predict.

Due to the limited amount of dataset used to train, the model only has an accurracy score of 70% and it is very biased on certain preferences . For example, it associates cool colour to Male etc

Here is a demo of the app.

Check out the app using this link

Further retraining will be done in the future as more data is accumulated.