Question

I need to create the "nnclassifier using the euclidean distance formula to find nearest neighbor. I am using the inis dataset from seabom. I have loaded the data and split the database into values and species. I have created a function to find the nearest neighbor by also calculating the euclidean distance and appending them to a distance list. I am able to see the positions of the nearest neighbors. Smart numpy as Import pandas as pd Import seaborn as sns import matplatsb.pyplot as plt Import scipy.stats as stats from sklearn.datasets import load breast cancer Iris - sns.load dataset(iris) Iris.shape, iris. species.shape . X contains the first four columns of the dataset attributes X - Iris. lloc , 3-1).values y contains the Labels of the columns y - Iris. lloc , 4).values train test split from sklearn.model_selection import train test split this will split the dataset into se trata adin 20% test data X train, X test, y train, y test train test split(x, ytest size-0.20) X train, shape, x_test. shape, y train.shape, y test.shape der top_neighbors (train instance 1. train Instance 2. k-5): distances - for idx in golon(train instance 1)): #calculate the distance between the two points by np. Linalg.non(x,y) distances.append(np. linalg.normtrain Instance 1 Sdx ] - train instance 2) return the top neighbors return np.argsort(distances) [ER] Idxs-top_nelghbors_traint.). X_teste.:]) y_trainides) I need to implement they pred which will allow my model to predict the flower. Please feel free to modify above code to more readability and functionality

Answer 1

import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import scipy.stats as stats
from sklearn.datasets import load_breast_cancer

In [1]:

iris=sns.load_dataset('iris')

In [3]:

iris.shape,iris.species.shape

Out[3]:

((150, 5), (150,))

In [4]:

X=iris.iloc[:,:-1].values

In [5]:

y=iris.iloc[:,4].values

In [6]:

from sklearn.model_selection import train_test_split

In [7]:

X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.20)

In [8]:

X_train.shape,X_test.shape,y_train.shape,y_test.shape

Out[8]:

((120, 4), (30, 4), (120,), (30,))

In [9]:

def top_neighbors(train_instance_1,train_instance_2,k=5):
    distances=[]
    for idx in range(len(train_instance_1)):
        distances.append(np.linalg.norm(train_instance_1[idx,:]-train_instance_2))
    return np.argsort(distances)[:k]

In [11]:

idxs=top_neighbors(X_train[:,:],X_test[0,:])
idxs
                

Out[11]:

array([ 66,  52,  68, 116, 106], dtype=int32)

In [12]:

y_train[idxs]

Out[12]:

array(['versicolor', 'versicolor', 'versicolor', 'versicolor',
       'versicolor'], dtype=object)

In [13]:

from sklearn.neighbors import KNeighborsClassifier

In [14]:

classifier=KNeighborsClassifier(n_neighbors=5)

In [15]:

classifier.fit(X_train,y_train)

Out[15]:

KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',
                     metric_params=None, n_jobs=None, n_neighbors=5, p=2,
                     weights='uniform')

In [20]:

y_pred=classifier.predict(X_test)
y_pred

Out[20]:

array(['versicolor', 'setosa', 'virginica', 'setosa', 'virginica',
       'versicolor', 'versicolor', 'virginica', 'versicolor', 'virginica',
       'versicolor', 'versicolor', 'setosa', 'virginica', 'setosa',
       'setosa', 'versicolor', 'setosa', 'versicolor', 'virginica',
       'versicolor', 'setosa', 'setosa', 'virginica', 'virginica',
       'versicolor', 'setosa', 'setosa', 'versicolor', 'versicolor'],
      dtype=object)

In [18]: