kaggle-titanic

About

This repository contains the code that achieves 76% score in the Kaggle Titanic Competition The code was developed to serve as an initial exploration of Scikit-Learn.

Pre-Processing

To achieve the result, the following were the pre-processing techniques applied:

• [Age], [Fare]: Missing Values were handled with Mean
• [Embarked]: Missing values were handled with Most Frequent Strategy
• [Embarked]: One Hot Encoder applied
• [Sex]: Ordinal Encoder applied
• [Age], [SibSp], [Fare], [Pclass]: Numerical Features that were normalized with MinMaxScaler

Model

The model built in this implementation is a simple Logistic Regression Model