Titanic Analysis

In this project, I analyzed the Titanic dataset and prediced the surival rates of the passengers on-board.

Exploratory Data Analysis

Here is a general breakdown:

Pre-Processing

• drop values that won't be useful to analysis
• add a variable in Sex where 0 is female, 1 is male
• P class variables
• Create random ages based on avg, std, and null freq
• Fill null spots in Age
• fill null values in Embarked
• fill null spot in Fare

Classifier Implementation

• features used to train: Pclass, Fare, male or not, and Age
• label used to train: Survived
• Used MLPClassifier

Model Validation

1.00 is the best possible ROC AUC, higher is better.