It rains today, will it rain again tomorrow?

Predicting real-time and accurate rainfall remains challenging for many decades due to its non-linear nature. Weather forecast can be divided into various terms such as real-time, short-term, middle-term, long-term. In this project, I focus on short term rainfall prediction whose forecasting range is 0-72 hours and able to predict next day rain. This research evaluates the effectiveness of different outlier detection, classification, and machine learning methods to predict rainfall in terms of weather forecast. There are various factors such as Humidity and Cloud are influencing the weather patterns and affects different parts of the world. My purpose is to find out those factors, which can play an important part in rainfall prediction

Technologies:
Dataset: weatherAUS.csv
Data Preprocessing: Handling missing values, feature Selection, handling data Correlation.
Feature Selection: Cutoff point for highly correlated variables
Outliers Detections: Clustering(K-Mean, DBScan & partition clustering)
Machine Learning Models: Logistic regression, Artificial Neural Network, Support Vector Machine
Performance Analysis: Confusion Mattrix & Roc Curve

For more information, please go through the final report and feel free to create an issue for any misunderstading.