🚢 Titanic Data Analytics

An exploratory and predictive analysis project using the Kaggle Titanic: Machine Learning from Disaster dataset, implemented in a Jupyter Notebook.

---

🧠 Overview

This project explores and models passenger survival using data from the RMS Titanic disaster. It covers:

• Data cleaning and exploratory data analysis (EDA)
• Feature engineering, visualization, and hypothesis testing
• Building and evaluating machine learning models to predict survival

---

🧩 Objectives

• Identify key factors influencing survival (e.g., age, sex, class, embarkation)
• Visualize patterns across demographic and socio-economic variables
• Train predictive models to classify survival outcomes
• Statistically validate relationships using tests such as Chi-square

---

🚀 Tech Stack

• Python libraries: pandas, numpy, matplotlib, seaborn, scipy
• Modeling frameworks: classifier algorithms like Random Forest, Logistic Regression
• Jupyter Notebook (PROJECT.ipynb)

---

📂 Repository Structure

Titanic_Data_Analytics/
│
├── PROJECT.ipynb         # Main notebook with full code and narrative
├── README.md             # This document
└── data/                 # Data folder (optional; dataset imported from Kaggle)

---

✅ Highlights

• Cleaning and preprocessing: Address missing values, data types, and outliers
• Exploratory plots: Analyze variables like age, sex, class, and survival via visualizations
• Statistical testing: Use chi-square tests to assess relationships (e.g., age groups vs survival)
• Modeling: Train and evaluate classification models on passenger data

---

📊 Insights & Results

• First-class passengers and female passengers had notably higher survival rates.
• Statistical tests confirmed age and passenger class significantly affect survival probability.
• Visualizations illustrated demographic patterns clearly.
• Predictive models achieved competitive accuracy in classifying survival.

---

📈 Future Enhancements

• Feature engineering with additional derived variables (e.g., titles from names, family size)
• Model tuning and ensembling for improved predictions
• Deployment as a web app for interactive user input and prediction

---

🧾 Credits & References

• Kaggle Titanic: Machine Learning from Disaster dataset
• Tutorials and walkthroughs on Titanic EDA and ML techniques (e.g., Analytics Vidhya, DataQuest)

---

👤 Author

Shashwat Srivastava
BCA Student – SRM University
GitHub: Shashwat970
Notebook: PROJECT.ipynb