🚢 Titanic Survival Prediction: Missing Values & Feature Engineering

A comprehensive machine learning project demonstrating production-ready data preprocessing pipelines with scikit-learn

Key Features • Quick Start • Methodology • Results • Contributing

---

📖 Overview

This project demonstrates a complete end-to-end machine learning pipeline for predicting passenger survival on the Titanic using the famous dataset. The focus is on building robust, reusable preprocessing pipelines that handle missing values and feature transformations professionally.

🎯 What You'll Learn

• ✅ Missing Value Detection — Identify and analyze patterns in incomplete data
• ✅ Imputation Strategies — Apply median, mode, and custom imputation techniques
• ✅ Feature Engineering — Transform raw features into ML-ready representations
• ✅ Pipeline Architecture — Build modular, production-ready sklearn pipelines
• ✅ Data Leakage Prevention — Implement proper train/test splitting strategies
• ✅ Model Evaluation — Assess classifier performance with comprehensive metrics

---

📊 Dataset

The Titanic Dataset contains information about 891 passengers from the RMS Titanic. The goal is to predict whether a passenger survived the sinking based on various features.

Feature Descriptions

Feature	Type	Description	Missing Values
pclass	Categorical	Passenger class (1st, 2nd, 3rd)	0
sex	Categorical	Gender (male/female)	0
age	Numerical	Age in years	177
sibsp	Numerical	# of siblings/spouses aboard	0
parch	Numerical	# of parents/children aboard	0
fare	Numerical	Ticket price (British Pounds)	0
embarked	Categorical	Port of embarkation (C/Q/S)	2
survived	Binary	Target variable (0 = No, 1 = Yes)	0

📈 Data Visualizations

Age Distribution	Embarked Distribution

---

🔧 Methodology

Pipeline Architecture

┌─────────────────────────────────────────────────────────────┐
│                    PREPROCESSING PIPELINE                    │
├─────────────────────────────────────────────────────────────┤
│                                                             │
│  ┌─────────────────┐         ┌─────────────────┐           │
│  │  NUMERICAL      │         │  CATEGORICAL    │           │
│  │  FEATURES       │         │  FEATURES       │           │
│  │                 │         │                 │           │
│  │  • age          │         │  • pclass       │           │
│  │  • sibsp        │         │  • sex          │           │
│  │  • parch        │         │  • embarked     │           │
│  │  • fare         │         │                 │           │
│  │                 │         │                 │           │
│  │  ┌───────────┐  │         │  ┌───────────┐  │           │
│  │  │ Imputer   │  │         │  │ Imputer   │  │           │
│  │  │ (median)  │  │         │  │ (mode)    │  │           │
│  │  └─────┬─────┘  │         │  └─────┬─────┘  │           │
│  │        │        │         │        │        │           │
│  │  ┌─────▼─────┐  │         │  ┌─────▼─────┐  │           │
│  │  │ Standard  │  │         │  │ OneHot    │  │           │
│  │  │ Scaler    │  │         │  │ Encoder   │  │           │
│  │  └───────────┘  │         │  └───────────┘  │           │
│  └────────┬────────┘         └────────┬────────┘           │
│           │                           │                     │
│           └───────────┬───────────────┘                     │
│                       │                                     │
│              ┌────────▼────────┐                           │
│              │ ColumnTransformer│                           │
│              └────────┬────────┘                           │
│                       │                                     │
├───────────────────────┼─────────────────────────────────────┤
│              ┌────────▼────────┐                           │
│              │   Logistic      │                           │
│              │   Regression    │                           │
│              └─────────────────┘                           │
│                    MODEL                                    │
└─────────────────────────────────────────────────────────────┘

Key Implementation Details

1. Numerical Feature Pipeline

numerical_pipeline = Pipeline([
    ('imputer', SimpleImputer(strategy='median')),  # Robust to outliers
    ('scaler', StandardScaler())                    # Zero mean, unit variance
])

2. Categorical Feature Pipeline

categorical_pipeline = Pipeline([
    ('imputer', SimpleImputer(strategy='most_frequent')),  # Mode imputation
    ('encoder', OneHotEncoder(handle_unknown='ignore'))     # Sparse encoding
])

3. Combined Preprocessor

preprocessor = ColumnTransformer([
    ('num', numerical_pipeline, numerical_features),
    ('cat', categorical_pipeline, categorical_features)
])

4. Full ML Pipeline

model = Pipeline([
    ('preprocessor', preprocessor),
    ('classifier', LogisticRegression(max_iter=100000))
])

---

📈 Results

Model Performance

Metric	Score
Accuracy	79.4%
Precision (Survived)	77%
Recall (Survived)	70%
F1-Score (Survived)	73%

Classification Report

              precision    recall  f1-score   support

           0       0.81      0.86      0.83       133
           1       0.77      0.70      0.73        90

    accuracy                           0.79       223
   macro avg       0.79      0.78      0.78       223
weighted avg       0.79      0.79      0.79       223

---

🚀 Quick Start

Prerequisites

• Python 3.8+
• pip package manager

Installation

# Clone the repository
git clone https://github.com/yourusername/featureEngineering.git
cd featureEngineering

# Create virtual environment (recommended)
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install pandas scikit-learn seaborn matplotlib numpy jupyter

Run the Notebook

jupyter notebook missing_values.ipynb

---

📁 Project Structure

featureEngineering/
├── 📓 missing_values.ipynb    # Main notebook with complete analysis
├── 📖 README.md               # Project documentation
├── 📊 age_plot.png            # Age distribution visualization
├── 📊 embarked_plot.png       # Embarked distribution visualization
├── 📁 .venv/                  # Virtual environment (not tracked)
└── 📁 .vscode/                # VS Code settings

---

🛠️ Technologies Used

Technology	Purpose
🐼 Pandas	Data manipulation and analysis
📊 Seaborn	Statistical data visualization
🤖 Scikit-learn	ML algorithms and preprocessing
📈 Matplotlib	Plotting and visualization
🔢 NumPy	Numerical computations

---

🎓 Learning Outcomes

By studying this project, you will learn:

Topic	Description
Missing Data Handling	Strategies for imputing incomplete datasets
Pipeline Architecture	Building modular, reusable preprocessing workflows
Data Leakage Prevention	Proper train/test splitting to avoid bias
Feature Engineering	Transforming raw data for ML consumption
Model Evaluation	Comprehensive metrics for classification tasks
Code Organization	Clean, documented, production-ready code

---

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

---

👤 Author

Ahmed Essam
Machine Learning & Data Science Enthusiast

---

📄 License

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

---

⭐ If you found this project helpful, please give it a star!

Made with ❤️ by Ahmed Essam