Starting With Random Forests

Made a start today also with Random Forests.

Still lots to learn - the model below scored only 0.73206, which is well below my best score for logistic regression.

 r <- randomForest(survived ~ male + pclass + fare + fare_per_person + 
age_class.interaction + sex_class + combined_age + family + age_squared + 
age_class_squared, data = train, importance = TRUE, ntree = 1000, do.trace = 100)
 
 ntree      OOB      1      2
  100:  18.18% 13.11% 26.32%
  200:  18.29% 13.30% 26.32%
  300:  18.07% 13.11% 26.02%
  400:  17.85% 12.57% 26.32%
  500:  17.96% 12.57% 26.61%
  600:  17.85% 12.39% 26.61%
  700:  18.18% 12.75% 26.90%
  800:  18.07% 12.57% 26.90%
  900:  18.18% 12.75% 26.90%
 1000:  18.07% 12.57% 26.90%