Kaggle Competition Entry to Forecast Rain in Australia using a Neural Network
Competition Link: https://www.kaggle.com/jsphyg/weather-dataset-rattle-package
This project was a binary classification problem using 10 years of historic data in Australia to predict whether it would rain tomorrow in each location based on the current days data. I used a neural network classifier to perform the classification and the final results were an accuracy of 84.03% and an F1 of 0.60 which is amongst some of the top performing models submitted.