Scribd
Upload
2 views

Predicting Customer Churn in Telecom

The document discusses customer churn in the telecommunications sector, defining it as the discontinuation of service by customers. It identifies common causes of churn, such as poor network quality, high pricing, and inadequate customer service, and presents a dataset from Kaggle containing 7043 customer records for analysis. The document also outlines data cleaning and visualization techniques used to analyze factors influencing churn.

Uploaded by

lavanyan1205
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
2 views

Predicting Customer Churn in Telecom

The document discusses customer churn in the telecommunications sector, defining it as the discontinuation of service by customers. It identifies common causes of churn, such as poor network quality, high pricing, and inadequate customer service, and presents a dataset from Kaggle containing 7043 customer records for analysis. The document also outlines data cleaning and visualization techniques used to analyze factors influencing churn.

Uploaded by

lavanyan1205
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 26

Telecommun

ications
Churn
Prediction
PREDICTING CUSTOMER CHURN IN TELECOM :

MISSING VALUES, OUTLIERS IN BILLING AMOUNTS,


INCONSISTENCIES IN CATEGORICAL VARIABLES.
What is
Customer
Customer churn is defined as when customers or subscribers
discontinue using sevices of firm.

Churn?
what is Telcom
Customer
Churn ?
Customers in telecmunication can switch to one Service privider to
another this can cause the telcom cutomer churn.
Common
Causes of
Telecom
Churn
Poor Network Quality (dropped calls, slow internet)

High Pricing (competitors offering better deals)

Poor Customer Service (unresolved complaints)

Lack of Personalization (customers feel undervalued)

Billing Issues (hidden charges, incorrect bills)


Here Telcom
Customer Churn
dataset is got
Loading Libraries and

from Kaggle
Data:

This data set is about 7043 cutomer records out of then 1869
Cutomers
Columns
Discription
lets see random 5
records of cutomer
tel.sample(
5)

tel.info
()
tel['TotalCharges']=pd.to_numeric(tel['TotalCharges'],errors='c
oerce')
tel.info()
temp=tel.copy()

tel['SeniorCitizen']=list(tel['SeniorCitizen'].map({0:'No',
1:'Yes'}))

tel['SeniorCitizen']
tel.info
()
Data Cleaning
tel[tel['TotalCharges'].isnu
ll()]
tel[tel['tenure']=
=0]
tel[tel['Dependents']=='Y
es']
# tel=tel[tel['tenure']!=0]
tel[tel['Contract']=='Two year']
['tenure'].describe()
df_temp=tel[(tel['Contract']=='Two year') &
(tel['PaymentMethod']=='Mailed check')]
# df_temp.isnull().sum
df_temp.dropna(inplace=Tr
ue)
df_temp.head(
.
2)
.

Cleaning Done
Data
Visulization
import seaborn as sns

import matplotlib.pyplot as
plt
plt.pie(tel['Churn'].value_counts(),labels=['No',
'Yes'],autopct='%1.1f%
%',colors=['green','red'],explode=[0.1, 0])

plt.title("Churn Percentage")

plt.show()
Data
Visulization
sns.countplot(data=tel,x='Ch
urn')
Data
Visulization
sns.countplot(data=tel,x='Churn',hue='ge
nder')
Data
Visulization
sns.countplot(data=tel,x='Contract',hue='Chur
n',palette="Set2")

plt.title('Churn Based on Contract Type')


Data
Visulization
df=tel[tel['Contract']=='Month-to-
month']

# sns.histplot(data=tel,tel['

df.sample()

sns.histplot(data=tel,x=df['MonthlyChar
ges'],hue='Churn')
Data
Visulization
sns.countplot(data=tel,x='Contract',hue='Ch
urn',palette="Set2")

plt.title('Churn Based on Contract Type')


Data
Visulization
df=tel[tel['Contract']=='Month-to-month']

# sns.histplot(data=tel,tel['

df.sample()

sns.histplot(data=tel,x=df['MonthlyCharges']
,hue='Churn')
Data
Visulization
sns.kdeplot(data=tel,x=df['MonthlyCharges']
,hue='Churn')
Data
Visulization
tel.sample()

sns.countplot(data=tel,x=tel['Dependents'],
hue=tel['Churn'],palette='Set1')
Data
Visulization
sns.countplot(data=tel,x=tel['PhoneService']
,hue='Churn',palette='Set1')
Data
Visulization
sns.countplot(data=tel,x=tel['InternetServic
e'],hue='Churn')

sns.countplot(data=tel,x=tel['OnlineSecurity
'],hue='Churn')

sns.countplot(data=tel,x=tel['DeviceProtecti
on'],hue=tel['Churn'])

sns.countplot(data=tel,x=tel['PaperlessBillin
g'],hue=tel['Churn'])
Data
Visulization
sns.countplot(data=tel,x=tel['TechSupport'],
hue=tel['Churn'])

sns.countplot(data=tel,x=tel['StreamingTV'],
hue=tel['Churn'])

sns.countplot(data=tel,x=tel['DeviceProtecti
on'],hue=tel['Churn'])

sns.countplot(data=tel,x=tel['StreamingMovi
es'],hue=tel['Churn'])
Data
Visulization
sns.kdeplot(data=tel,x=tel['tenure'],hue=tel
['Churn'])
Data
Visulization
tdf=tel[['tenure','MonthlyCharges','TotalChar
ges']]
tcorrelational=tdf.corr()

tcorrelational
CONCLU
SION
thanyo
k u!
PREDICTING CUSTOMER CHURN IN
TELECOM

You might also like