AI Features

Pre-Processing

Learn how we can calculate the ticket class modifier, the gender modifier, and pre-processing.

The pre-processing covers the calculation of the modifiers. We start with the ticket class.

Calculating the ticket class modifier

Javascript (babel-node)
# get the modifier given the passenger's pclass
def get_modifier_pclass(pclass):
# number of passengers with the same pclass
cnt_surv_pclass = len(survivors[survivors.Pclass.eq(pclass)])
# backward probability
p_cl_surv = cnt_surv_pclass/cnt_survivors
# probability of the evidence
p_cl = len(train[train.Pclass.eq(pclass)])/cnt_all
return p_cl_surv/p_cl

We define a function that takes the passenger’s pclass as input. The Pclass column in our dataset is the ticket class (1 = 1st, 2 = 2nd, 3 = 3rd).

We calculate the backward probability P(PclassSurvived)P(Pclass∣Survived) by dividing the passengers who survived having the given ticket class (cnt_surv_pclass) in line 4 by all survivors (cnt_survivors) in line 7. Then, we calculate the probability of a passenger owning the given ticket class. The number of passengers with the given ticket class is divided by the total number of passengers in line 10.

The ...