Skip to content

Latest commit

 

History

History
87 lines (63 loc) · 2.77 KB

README.md

File metadata and controls

87 lines (63 loc) · 2.77 KB
Raw
# Interstellar Passenger Rescue - Spaceship Titanic

## Overview

Welcome to the Interstellar Passenger Rescue project! This repository contains the code and resources for predicting which passengers from the Spaceship Titanic were transported to an alternate dimension after encountering a spacetime anomaly.

This project utilizes machine learning techniques, specifically TensorFlow Decision Forests, to analyze passenger data and predict the outcomes.

## Dataset

The dataset consists of the following files:
- `train.csv`: Training dataset containing information about passengers.
- `test.csv`: Test dataset for making predictions.
- `sample_submission.csv`: Sample submission format for Kaggle competition.

## Requirements

To run the code in this repository, ensure you have the following installed:
- Python 3
- TensorFlow 2.x
- TensorFlow Decision Forests
- Pandas
- NumPy
- Seaborn
- Matplotlib

Install dependencies using:
```bash
pip install -r requirements.txt

Getting Started

  1. Clone this repository:

    git clone https://github.com/shubvats/Interstellar-Passenger-Rescue-Spaceship-Titanic.git
cd Interstellar-Passenger-Rescue-Spaceship-Titanic

  2. Download the dataset from Kaggle or use the provided train.csv and test.csv.

  3. Explore the notebooks and scripts:

    • Spaceship_Titanic_Prediction.ipynb: Jupyter notebook containing the entire workflow from data preprocessing to model evaluation and prediction.

  4. Train and evaluate the model:

    • Run the notebook or scripts to train the TensorFlow Decision Forests model on the training dataset.
    • Evaluate the model performance on the validation dataset.

  5. Make predictions:

    • Use the trained model to make predictions on the test dataset (test.csv).
    • Generate a submission file (submission.csv) in the format required for the Kaggle competition.

Model Evaluation

The trained model achieves an accuracy of approximately 81.41% on the validation dataset, demonstrating effective prediction capabilities.

Folder Structure

├── data/
│   ├── train.csv
│   ├── test.csv
│   ├── sample_submission.csv
├── notebooks/
│   ├── Spaceship_Titanic_Prediction.ipynb
├── models/
│   ├── trained_model/
│   │   ├── saved_model.pb
│   │   ├── variables/
├── README.md
├── requirements.txt
└── submission.csv

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

  • Kaggle "Spaceship Titanic" competition for providing the dataset and challenge.
  • TensorFlow Decision Forests community for the powerful machine learning tools.

Feel free to contribute, provide feedback, or raise issues if you encounter any problems.