Global Weather Trend Forecasting and Climate Pattern Analysis

Overview

This project analyzes the Global Weather Repository dataset to forecast weather trends and explore climate, air quality, anomaly, and spatial patterns across global cities.

PM Accelerator Mission

Product Manager Accelerator is designed to support product management professionals through every stage of their careers, helping aspiring and experienced PMs build skills, prepare for interviews, and transition into product roles.

Dataset

Dataset: Global Weather Repository Source: Kaggle World Weather Repository

Methods

• Data cleaning and preprocessing
• Missing value handling
• Exploratory data analysis
• Temperature and precipitation visualization
• Correlation analysis
• Anomaly detection using Isolation Forest
• Forecasting using multiple machine learning models
• Ensemble forecasting
• Feature importance using permutation importance
• Air quality and environmental impact analysis
• Spatial and geographical weather pattern analysis

Models

• Linear Regression
• Ridge Regression
• Random Forest Regressor
• Gradient Boosting Regressor
• Ensemble Model

Evaluation Metrics

• MAE
• RMSE
• R2 Score

Outputs

• model_comparison.csv
• feature_importance.csv
• country_weather_summary.csv
• spatial_weather_summary.csv
• weather_trend_forecasting_report.md

How to Run

1. Upload GlobalWeatherRepository.csv to Google Drive.
2. Open the notebook in Google Colab.
3. Mount Google Drive.
4. Run all cells from top to bottom.
5. Outputs will be saved in the folder weather_trend_forecasting_outputs.

Author

Divya Natekar