Weather Phenomenon Recognition using Deep Learning

This project develops deep learning models to classify weather phenomena from images.

Project Overview

Part of MIS373 - Advanced AI For Business course. Uses Python and Jupyter Notebook to analyze and classify 2,726 weather images into 5 categories: fogsmog, rain, rime, sandstorm, and snow.

Key Components

• Data exploration and preprocessing
• Development of various deep learning models
• Performance comparison and analysis
• Discussion on improving model performance for real-world applications

Tools and Technologies

• Python
• Jupyter Notebook
• Google Colab (with GPU)
• Libraries: numpy, pandas, tensorflow, etc. and applying deep learning frameworks

Results

The result has been summarized directly in the Andrew_Nguyen_MIS373A2_Task2.ipynb file. The MIS373_Task2_Testing.ipynb file are where I first training multiple models to get the best one

Author

Andrew Nguyen

Acknowledgements

• Dataset provided by Jehan Bhathena on Kaggle
• Project developed for MIS373 at Deakin University