Titanic Passenger Age Prediction

This project is a machine learning exploration focused on predicting the age of passengers from the Titanic dataset. It serves as a practical exercise in data cleaning, feature engineering, and model evaluation.

This repository documents the end-to-end process, from the data preprocessing to the evaluation of model performance.

---

Dataset

The project utilizes the tested.csv file from the Titanic dataset, which contains various details for each passenger, including their class, fare, and family size.

---

🔧 How to Run the Project

1. Clone the repository:

   git clone [https://github.com/AS200585/Titanic-ML-Project.git](https://github.com/AS200585/Titanic-ML-Project.git)
   cd Titanic-ML-Project

2. Install the required libraries:

   pip install pandas numpy scikit-learn matplotlib

3. Run the script:

   python titanic-prediction.py

---

Results

An initial set of models was built using the available features. This approach produced extremely high R-squared scores, suggesting near-perfect prediction capabilities.

Model	Test R² Score
Linear Regression	0.830
Lasso Regression	0.844
Gradient Boosting	0.999

Analysis: Uncovering Data Leakage

Upon review, it was discovered that the near-perfect scores of the GradientBoostingRegressor were a result of data leakage.

---

🚀 Future Improvements

The following steps are planned to improve the model's predictive power:

1. Smarter Imputation: Replace the simple fillna(0) strategy for missing Age values with a more robust method, such as filling with the median age.
2. Feature Engineering: Convert categorical columns like Sex and Embarked into numerical format using one-hot encoding.
3. Feature Selection: Remove non-informative features like PassengerId that only add noise to the model.
4. Hyperparameter Tuning: Once the features are improved, fine-tune model parameters to optimize performance and reduce overfitting.