This pull requests adds a simple binary classifier. The idea is to feed it content and labels for Female or Male depending on the websites in order to build the classifier. In the end we can count the frequency in the result of the classified test data (which will be an array of 0s and 1s) and assign a probability for the user to be either male or female.
Eg. gives 5 websites and their content the result of the classifier will be [1,1,0,0,0] with 1 = Male0 = Female so we are 60% certain the user is Female and 40% that it could be Male
tiriplicamihai
commented
8 years ago
This pull requests adds a simple binary classifier. The idea is to feed it content and labels for Female or Male depending on the websites in order to build the classifier. In the end we can count the frequency in the result of the classified test data (which will be an array of 0s and 1s) and assign a probability for the user to be either male or female.
Eg. gives 5 websites and their content the result of the classifier will be
[1,1,0,0,0]with1 = Male0 = Femaleso we are 60% certain the user is Female and 40% that it could be Male