Emotion Classification - Tech Fest AI Hackathon

Welcome to the Emotion Classification project! This repository contains my solution for the Emotional Fluctuations Classification challenge from the Tech Fest 1.0 AI Hackathon, hosted at my university and featured on Kaggle.

🚀 Project Overview

The goal of this project is to predict emotional fluctuation levels based on various features using machine learning. I built a robust pipeline using scikit-learn, achieving a strong validation accuracy of 0.97 with a Random Forest Classifier.

🧑‍💻 What’s Inside

• Jupyter Notebook: Step-by-step code, explanations, and results.
• Data Preprocessing: Handling missing values, scaling, and encoding features.
• Model Building: Random Forest Classifier with a full ML pipeline.
• Evaluation: Validation accuracy and prediction output.
• Submission File: Automatically generated for Kaggle format.

📊 Key Features

• Clean and reproducible ML pipeline using Pipeline and ColumnTransformer.
• Separate preprocessing for numerical and categorical features.
• Easy-to-follow notebook for learning and experimentation.
• Ready for adaptation to similar classification problems.

🗂️ Repository Structure

• notebook.ipynb — Main notebook with all code and results.
• train.csv / test.csv — Example data files (add if allowed).
• submission.csv — Output file generated by the notebook.

🏁 Getting Started

1. Clone the repository:

   git clone https://github.com/Mahadasghar/Emotion_classifier.git

2. Add the dataset files (train.csv, test.csv) to the project folder.
3. Open and run notebook.ipynb in Jupyter or VS Code.
4. Check your results and experiment with the pipeline!

💡 Why Share This?

Although I missed the official competition deadline, this project demonstrates my skills in data science, feature engineering, and model building. I hope it helps others learn and inspires improvements!

📬 Contact

Feel free to reach out via GitHub Issues for questions or suggestions.