Question

Split it into training and test data. Apply logistic regression and KNN algorithm. You can use R or Python to implement them. Which one of the algorithm fares better?

The dataset can be found at https://www.kaggle.com/uciml/indian-liver-patient-records/version/1

Much much thx for help

Answer 1

The python 3 code for KNN is given below. Make sure you have libraries like numpy, pandas and sklearn installed in python.

******************KNN_liver_model.py*******************************

import numpy as np
from sklearn import neighbors,model_selection

import pandas as pd
from math import nan

#load the dataset
df=pd.read_csv('indian_liver_patient.csv')

#replace NaN or empty cells with 0.5
df.replace(nan, 0.5, inplace=True)
#encode Male as 1 and Female as 2
df.replace({'Male':1.0,'Female':2.0},inplace=True)

#X contains the features and Y contains the final classification
X=np.array(df.drop(['Dataset'],1))
y=np.array(df['Dataset'])

#keep 0.1% as the test data
X_train,X_test,y_train,y_test=model_selection.train_test_split(X,y,test_size=0.1)

clf=neighbors.KNeighborsClassifier(n_neighbors=4);
clf.fit(X_train,y_train)
accuracy=clf.score(X_test,y_test)

print('The accuracy is:'+str(accuracy))