- Introduction In this project, we worked on the famous Titanic dataset to analyze passenger data and predict whether a passenger survived or not. The goal was to understand how different factors like gender, class, and fare influenced survival, and then build a machine learning model to make predictions.
- Objective The main objective of this project was to: • Explore and understand the dataset • Clean and preprocess the data • Visualize important patterns • Build a machine learning model to predict survival
- Dataset Description The dataset contains information about passengers such as: • Age • Gender (Sex) • Passenger Class (Pclass) • Fare • Number of family members (SibSp, Parch) • Embarked location These features help in analyzing survival patterns.
- Data Preprocessing Before building the model, the data needed to be cleaned and prepared: • Missing values in Age were filled using median values • Missing values in Embarked were filled using mode • Missing values in Deck were handled using a placeholder ("Unknown") • Unnecessary columns such as who, alive, deck, etc., were removed • Categorical values like Sex, Embarked, and Alone were converted into numerical form This step ensured the dataset was ready for machine learning.
- Data Visualization & Insights Several graphs were plotted to understand the data better: • Survival Distribution: Showed how many passengers survived vs not survived • Survival by Gender: Females had a significantly higher survival rate than males • Fare vs Class (Box Plot): First-class passengers paid higher fares Key Insights: • Female passengers had a much higher chance of survival • Passengers in higher classes (especially first class) survived more • Higher fare was associated with better survival chances These insights helped us understand patterns before building the model.
- Model Building We used Logistic Regression, a classification algorithm, since the output (survived or not) is binary. Steps followed: • Selected input features (X) and output (y = survived) • Split the dataset into training (80%) and testing (20%) • Trained the model on training data • Predicted results on test data
- Model Evaluation Accuracy The model achieved an accuracy of: 82.5% This means the model correctly predicted survival for about 82 out of 100 passengers.
Confusion Matrix Interpretation • 69 passengers were correctly predicted as not survived • 49 passengers were correctly predicted as survived • 13 passengers were incorrectly predicted as survived • 12 passengers were incorrectly predicted as not survived Overall, most predictions were correct, with relatively few errors.
- Conclusion This project shows that: • Gender and passenger class are strong factors affecting survival • Data visualization helps in understanding patterns clearly • Logistic Regression performs well for this classification problem Even though the model is not perfect, it gives good accuracy and meaningful insights.
- Learning Outcomes Through this project, we learned: • Data cleaning and preprocessing techniques • Data visualization using Seaborn and Matplotlib • Converting categorical data into numerical form • Building and evaluating a machine learning model • Interpreting results using accuracy and confusion matrix