In this final project, the goal is to build a machine-learning model and dashboard for the Titanic Survival Prediction Model based on some parameters. The Dashboard will be built in Streamlit and the model will be using random forest.
- Load the dataset and explore its structure.
- Clean the data by handling missing values, interpolating data, and removing outliers.
- Perform feature engineering if needed.
- Visualize the distribution of Titanic Passenger parameters/features.
- Understand the relationships between different features.
- Analyze statistical summaries, visualizations, and correlations.
- Use machine learning algorithms (e.g., Random Forest, XGBoost, KNN, Logistic Regression) to predict Passenger Survivability.
- Train the models on a training set and evaluate them on a validation set.
- Assess model performance using metrics such as accuracy, confusion matrices, and classification reports.
- Compare different algorithms to select the most suitable one.
- Create a Streamlit dashboard for an interactive user interface.
- Allow users to input new data, and the trained model predicts Passenger Survivability.
- Integrate visualizations, including count plots and heatmaps, into the Streamlit app.
- pandas: pandas
- sklearn: scikit-learn
- matplotlib: matplotlib
- seaborn: seaborn
- numpy: numpy
- Summarize findings and insights gained from the data.
- Discuss recommendations or further steps, such as refining the model, exploring additional features, or improving data quality.