This project builds and evaluates machine learning models to predict whether it will rain today in Melbourne, Australia. Using historical weather data and Scikit-learn pipelines, we compare the performance of two classifiers: RandomForestClassifier and LogisticRegression.
🔍 Project Highlights End-to-end ML pipeline with Pipeline and GridSearchCV
Data preprocessing using column transformers (handling numeric and categorical features)
Feature importance visualization for model interpretability
Evaluation with classification report and confusion matrix
Comparison based on:
✅ Accuracy
🔢 Number of correct predictions
📈 True Positive Rate (recall) for predicting rain
📊 Results Summary Accuracy (both models): 83%
Correct Predictions: 1,254 out of 1,512
True Positive Rate (Logistic Regression): 51%
📁 About the Dataset The original dataset was obtained from Kaggle, and originates from the Australian Government's Bureau of Meteorology. Column descriptions were gathered from the official climate data guide.
The dataset includes daily weather observations from 2008 to 2017 across various Australian locations.
- Kaggle :https://www.kaggle.com/datasets/jsphyg/weather-dataset-rattle-package/
- Australian Government's Bureau of Meteorology http://www.bom.gov.au/climate/dwo/
- Column definitions were gathered from http://www.bom.gov.au/climate/dwo/IDCJDW0000.shtml