Explain Naïve Bayes along with a case study in python

First we import the libraries and the datasets

import pandas as pd

import bumpy as np

from sklearn.naive_bayes import Gaussian NB

from sklearn.naive_bayes import MultinomialNB

from sklearn.model_selection import train_test_split

from sklearn.metrics import confusion_matrix

salary_train = pd.read_csv("SalaryData_Train.csv")

salary_test = pd.read_csv("SalaryData_Test.csv")

string_columns=["workclass","education","maritalstatus","occupation","relationship","race","sex","native"]

Now we perform some preprocessing of the data and split into feature and target variable

from sklearn import preprocessing

number = preprocessing.LabelEncoder()

for i in string_columns:

    salary_train[i] = number.fit_transform(salary_train[i])

    salary_test[i] = number.fit_transform(salary_test[i])

colnames = salary_train.columns

len(colnames[0:13])

trainX = salary_train[colnames[0:13]]

trainY = salary_train[colnames[13]]

testX = salary_test[colnames[0:13]]

testY = salary_test[colnames[13]]

Now we will fit and predict the model for Gaussian Naïve Bayes

sgnb = GaussianNB()

smnb = MultinomialNB()

spred_gnb = sgnb.fit(trainX,trainY).predict(testX)

confusion_matrix(testY,spred_gnb)

print ("Accuracy",(10759+1209)/(10759+601+2491+1209))

Again we will fit and predict the model for Multinomial Naïve Bayes

spred_mnb = smnb.fit(trainX,trainY).predict(testX)

confusion_matrix(testY,spred_mnb)

print("Accuracy",(10891+780)/(10891+780+2920+780))