Implementation of Naive Bayes classifier to classify a given name as either a girl's or boy's name.
Problem statement:
When a foreigner comes to the States, sometimes they got very confused about baby’s names to tell it is a boy’s name or a girl’s name, like Jack and Jane. For this question, you will implement a Naive Bayes classifier to classify if a given test baby’s name is a boy’s name or a girl’s name. In the dataset, you can find a boy’s name list, boy names.csv, which contains 1000 commonly used boy’s names. You can also find a girl’s name list, girl names.csv, which contains 1000 commonly used girl’s names. In order to implement a NB classifier, you need to:
• Compute P(X|Y )
• Compute P(Y ) but here we don’t need you to compute this as indicated in the hint
• Compute log probability log(P(Y =1|x)/P(Y =−1|x))
• decide the classification result
Questions: • Implement a NB classifier for the name problem by using the two training files boy names.csv and girl names.csv.
• Test your algorithm using the test dataset test names.csv. For now you can just output your classification result based on which class has higher probability based on your model. Export your result as a .csv file, using “-1” to indicate a boy’s name and “+1” to indicate a girl’s name.
