Would You have Survived the Titanic?

Have you ever wondered if you would have survived the Titanic? Now you can get online predictions - at https://iabrady.shinyapps.io/shiny

• Uses a machine learning algorithm for accurate results!
• Easy to use - instructions are on the screen
• Gives immediate feedback with your survival likelihood and probability

Titanic trainging data courtesy of https://www.kaggle.com/

The learning algorithm is fast and based on a boosting algorithm

• Gives immediate feedback with your survival likelihood and probability
• incorporates training, cross-validation and error minimisation

library(dplyr); library(gbm)
train <- read.csv("train.csv", stringsAsFactors=T)
tdata <- select(train, 2:3, 5, 6); tdata <- na.omit(tdata)
boost.data <- gbm(Survived~., data=tdata, distribution="gaussian", 
                  n.trees=323, interaction.depth=3, shrinkage=0.01, 
                  verbose=F, train.fraction=0.8, cv.folds=3)

There is a huge following for the Titanic. It has fascinated people for a century and still does. For a casual user:

• Easy to use - anyone can follow the steps
• Gives immediate greatificatin with feedback on survival likelihood and probability
• Users can specify their own information or imagine being someone else!

A nice simple app with great opportunity for the future Enhancements are planned:

• Improving the user interface
• adding other funtionality to give more summary information about Titanic passengers
• Including graphics and other interest coponents
  

See the app at https://iabrady.shinyapps.io/shiny