Question

7.3 Consider a set of three records with two numeric fields X, and Y. Type is the categorical target variable. Record X 3.0 3.0 3.0 3.0 Small Large Large (a) Suppose we have a new record (D) with values X-1.5 and Y-3.0. Use k-NN simple unweighted voting (for different values of k-1,2, and 3) to determine the classification of D. A scatter plot of Y against X shows the location of the three training records, and new unclassified record. 3.5 2.5 0.5 X values (b) Suppose your data set is altered with two small changes: the target class for Record A is changed to "Large", and the target class of Record C is changed to "Small". Use k-NN simple unweighted voting with different values of k (1,2,3) to classify the new record (D) with values X-1.5 and Y-3.0 Record X 3.0 3.0 3.0 3.0 Large Large Small (c) Refer to 7.3b. Find the Euclidean distance between each pair of points. Using these points, verify that Euclidean distance is a true distance metric

Answer 1

a.

import pandas as pd

import numpy as np

from sklearn.neighbors import KNeighborsClassifier

t = pd.DataFrame({"l" : [3,3,1], "b" : [1,3,3], "op":['S',"L","L"] })

X = t.iloc[:,0:-1]

Y = t.iloc[:,-1]

for k in (1,2,3):

    knn = KNeighborsClassifier(n_neighbors=k).fit(X,Y)

    print("predicted size is" ,knn.predict([[1.5,3]])[0])

predicted size is L
predicted size is L
predicted size is L

b.

import pandas as pd

import numpy as np

from sklearn.neighbors import KNeighborsClassifier

t = pd.DataFrame({"l" : [3,3,1], "b" : [1,3,3], "op":['L',"S","S"] })

X = t.iloc[:,0:-1]

Y = t.iloc[:,-1]

for k in (1,2,3):

    knn = KNeighborsClassifier(n_neighbors=k).fit(X,Y)

    print("predicted size is" ,knn.predict([[1.5,3]])[0])

predicted size is S
predicted size is S
predicted size is S