7/17/23, 6:23 PM LP-III_KNN_SVM - Jupyter Notebook

In [1]: import pandas as pd

In [2]: df=pd.read_csv('C:/shubhangi/2023-24/LP-III_ML/Assignment 2/emails.csv')

In [3]: df.head()
​

Out[3]:
Email
the to ect and for of a you hou ... connevey jay valued lay infrastructure military
No.

Email
0 0 0 1 0 0 0 2 0 0 ... 0 0 0 0 0 0
1

Email
1 8 13 24 6 6 2 102 1 27 ... 0 0 0 0 0 0
2

Email
2 0 0 1 0 0 0 8 0 0 ... 0 0 0 0 0 0
3

Email
3 0 5 22 0 5 1 51 2 10 ... 0 0 0 0 0 0
4

Email
4 7 6 17 1 5 2 57 0 9 ... 0 0 0 0 0 0
5

5 rows × 3002 columns

In [4]: df.shape

Out[4]: (5172, 3002)

In [5]: #input data

x=df.drop(['Email No.','Prediction'],axis=1)
#output data
y=df['Prediction']

In [6]: x.shape

Out[6]: (5172, 3000)

In [7]: x.dtypes

Out[7]: the int64

to int64
ect int64
and int64
for int64
...
infrastructure int64
military int64
allowing int64
ff int64
dry int64
Length: 3000, dtype: object

localhost:8891/notebooks/LP-III_KNN_SVM.ipynb 1/6
7/17/23, 6:24 PM LP-III_KNN_SVM - Jupyter Notebook

In [8]: set(x.dtypes)

Out[8]: {dtype('int64')}

In [10]: import seaborn as sns

sns.countplot(x=y)

Out[10]: <Axes: xlabel='Prediction', ylabel='count'>

In [11]: y.value_counts()

Out[11]: 0 3672
1 1500
Name: Prediction, dtype: int64

In [14]: pip install scikit-learn

Requirement already satisfied: scikit-learn in c:\programdata\anaconda3\lib\site-packa

ges (1.2.2)
Requirement already satisfied: numpy>=1.17.3 in c:\programdata\anaconda3\lib\site-pack
ages (from scikit-learn) (1.24.3)
Requirement already satisfied: scipy>=1.3.2 in c:\programdata\anaconda3\lib\site-packa
ges (from scikit-learn) (1.10.1)
Requirement already satisfied: joblib>=1.1.1 in c:\programdata\anaconda3\lib\site-pack
ages (from scikit-learn) (1.2.0)
Requirement already satisfied: threadpoolctl>=2.0.0 in c:\programdata\anaconda3\lib\si
te-packages (from scikit-learn) (2.2.0)
Note: you may need to restart the kernel to use updated packages.

In [15]: from sklearn.preprocessing import MinMaxScaler

localhost:8891/notebooks/LP-III_KNN_SVM.ipynb 2/6
7/17/23, 6:24 PM LP-III_KNN_SVM - Jupyter Notebook

In [16]: scaler=MinMaxScaler()

In [17]: x_scaled=scaler.fit_transform(x)

In [18]: x_scaled

Out[18]: array([[0. , 0. , 0. , ..., 0. , 0. ,

0. ],
[0.03809524, 0.09848485, 0.06705539, ..., 0. , 0.00877193,
0. ],
[0. , 0. , 0. , ..., 0. , 0. ,
0. ],
...,
[0. , 0. , 0. , ..., 0. , 0. ,
0. ],
[0.00952381, 0.0530303 , 0. , ..., 0. , 0.00877193,
0. ],
[0.1047619 , 0.18181818, 0.01166181, ..., 0. , 0. ,
0. ]])

In [19]: #cross validation

from sklearn.model_selection import train_test_split

In [20]: x_train,x_test,y_train,y_test=train_test_split(x_scaled,y,random_state=0,test_size=0.25

In [22]: x_scaled.shape

Out[22]: (5172, 3000)

In [23]: x_train.shape

Out[23]: (3879, 3000)

In [24]: x_test.shape

Out[24]: (1293, 3000)

In [26]: from sklearn.neighbors import KNeighborsClassifier

In [27]: knn=KNeighborsClassifier(n_neighbors=5)

In [28]: knn.fit(x_train,y_train)

Out[28]: ▾ KNeighborsClassifier
KNeighborsClassifier()

In [29]: #predict on test data

y_pred=knn.predict(x_test)

In [31]: #import evaluation matrix

from sklearn.metrics import ConfusionMatrixDisplay,accuracy_score

localhost:8891/notebooks/LP-III_KNN_SVM.ipynb 3/6
7/17/23, 6:24 PM LP-III_KNN_SVM - Jupyter Notebook

In [32]: from sklearn.metrics import classification_report

In [33]: ConfusionMatrixDisplay.from_predictions(y_test,y_pred)

Out[33]: <sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x207c27b6c50>

In [34]: y_test.value_counts()

Out[34]: 0 929
1 364
Name: Prediction, dtype: int64

In [35]: accuracy_score(y_test,y_pred)

Out[35]: 0.871616395978345

In [36]: print(classification_report(y_test,y_pred))

precision recall f1-score support

0 0.98 0.84 0.90 929

1 0.70 0.95 0.81 364

accuracy 0.87 1293

macro avg 0.84 0.89 0.85 1293
weighted avg 0.90 0.87 0.88 1293

In [37]: import numpy as np

import matplotlib.pyplot as plt

localhost:8891/notebooks/LP-III_KNN_SVM.ipynb 4/6
7/17/23, 6:24 PM LP-III_KNN_SVM - Jupyter Notebook

In [40]: error = []
for k in range(1,41):
knn= KNeighborsClassifier(n_neighbors=k)
knn.fit(x_train,y_train)
pred=knn.predict(x_test)
error.append(np.mean(pred!=y_test))

In [39]: error

Out[39]: [0.10827532869296211,
0.10982211910286156,
0.12296983758700696,
0.11523588553750967,
0.12838360402165508,
0.1214230471771075,
0.15158546017014696,
0.14849187935034802,
0.17246713070378963,
0.16705336426914152,
0.1871616395978345,
0.18329466357308585,
0.21500386697602475,
0.21345707656612528,
0.22815158546017014,
0.2266047950502707,
0.23588553750966745,
0.23356535189481825,
0.2459396751740139,
0.24361948955916474,
0.2559938128383604,
0.2552204176334107,
0.2699149265274555,
0.2691415313225058,
0.2822892498066512,
0.28306264501160094,
0.2954369682907966,
0.2923433874709977,
0.3039443155452436,
0.300077339520495,
0.30549110595514306,
0.30549110595514306,
0.31245166279969067,
0.31245166279969067,
0.3194122196442382,
0.317092034029389,
0.32637277648878577,
0.32559938128383603,
0.33410672853828305,
0.3325599381283836]

In [41]: knn=KNeighborsClassifier(n_neighbors=1)

In [42]: knn.fit(x_train,y_train)

Out[42]: ▾ KNeighborsClassifier
KNeighborsClassifier(n_neighbors=1)

localhost:8891/notebooks/LP-III_KNN_SVM.ipynb 5/6
7/17/23, 6:24 PM LP-III_KNN_SVM - Jupyter Notebook

In [43]: y_pred=knn.predict(x_test)

In [44]: accuracy_score(y_test,y_pred)

Out[44]: 0.8917246713070379

In [45]: from sklearn.svm import SVC

In [*]: svm=SVC(kernel='linear')

In [*]: svm.fit(x_train,y_train)

In [*]: y_pred=svm.predict(x_test)

In [54]: accuracy_score(y_test,y_pred)

Out[54]: 0.9767981438515081

In [ ]: #linear 0.9767981438515081
​

localhost:8891/notebooks/LP-III_KNN_SVM.ipynb 6/6