6 April 2019

Motivation

An interesting, if somewhat classic, problem in machine learning is predicting whether a person would survive the Titanic disaster.

This application aims to answer this question: Given an age, sex and passenger class onboard the ship, would the passenger survive?

Demonstration

For example: James Alfred Brown is a male adult travelling in 1st class. Would he have survived?

The server's output is the following:

paste('Will the person survive on the Titanic?', pred)
## [1] "Will the person survive on the Titanic? No"

Survival Probabilities

The web application also presents a histogram of the survival probability for different passenger types, with the user able to customize the class for which the probability is shown.

Conclusions

  • Simple application for predicting whether a person would have survived the Titanic disaster
  • Informative and insightful for historians and relatives of those who were onboard
  • Please consider investing in my application!