A high accuracy Machine Learning project, implementing various models to accurately predict the survival of the on board passengers in Titanic. This project is binary classification problem, where the passenger either survived (1) or died (0).
Different models were used to compare and contrast their respective results and accuracy after having performed feature engineering on the data. The different models used were:
- Logistic Regression
- KNN
- Support Vector Machines
- Naive Bayes classifier
- Decision Tree
- Random Forrest
- Linear Discriminant Analysis
- Ada Boost Classifier
- Gradient Boosting Classifier
The technique of Hyper-Parameters Tuning was also applied to a few of the above models to further enhance the accuracy. The highest accuracy obtained was around 90% with the Random Forest classifier after appying the estimator obtained from parameter tuning of Random Forest.
All the further information regarding the project and dataset is available on kaggle. Link for data set: https://www.kaggle.com/c/titanic/data