Client Retention Dataset ClientID Phone Internet TVService MoviesService Gender SeniorCitizen Married HasDependents MonthsAsClient SecurityService BackupService DeviceProtection TechSupport ContractType Paperless PaymentMethod MonthlyCharges TotalCharges Quit PB 1 No DSL No No Female 0 Yes No 1 No Yes No No Monthly Yes e check 29 85 29 85 No PB 2 Yes DSL No No Male 0 No No 34 Yes No Yes No 1 Year No Check 56 95 1889 5 No PB 3 Yes DSL No No Male 0 No No 2 Yes Yes No No Monthly Yes Check 53 85 108 15 Yes PB 4 No DSL No No Male 0 No No 45 Yes No Yes Yes 1 Year No EBT 42 3 1840 75 No PB 5 Yes FiberOptic No No Female 0 No No 2 No No No No Monthly Yes e check 70 7 151 65 Yes PB 6 Yes FiberOptic Yes Yes Female 0 No No 8 No No Yes No Monthly Yes e check 99 65 820 5 Yes PB 7 Yes FiberOptic Yes No Male 0 No Yes 22 No Yes No No Monthly Yes Credit Card 89 1 1949 4 No PB 8 No DSL No No Female 0 No No 10 Yes No No No Monthly No Check 29 75 301 9 No PB 9 Yes FiberOptic Yes Yes Female 0 Yes No 28 No No Yes Yes Monthly Yes e check 104 8 3046 05 Yes PB 10 Yes DSL No No Male 0 No Yes 62 Yes Yes No No 1 Year No EBT 56 15 3487 95 No The dataset consists of 19 variables and the target variable Below is the description of all variables ClientID The internal ID of PythonBell clients Phone Does the client have phone line Yes, No Internet Does the client have internet DSL, FiberOptic, None TVService Does the client have TV Service Yes, No, NotApplicable MoviesService Does the client have Movie Service Yes, No, NotApplicable Gender The gender of the client Male, Female SeniorCitizen Is the client a senior citizen 1 Yes, 0 No Married Is the client married Yes, No HasDependents Does client have dependents Yes, No MonthsAsClient How many months the client has been a customer continuous variable SecurityService Does the client have the Security addon Yes, No, NotApplicable BackupService Does the client have the Backup addon Yes, No, NotApplicable DeviceProtection Does the client have the Protection addon Yes, No, NotApplicable TechSupport Does the client have the Support addon Yes, No, NotApplicable ContractType Contract length Monthly, 1 Year, 2 Years Paperless Bill sent electronically Yes, No PaymentMethod The method the client pays e check, Check, EBT, Credit Card MonthlyCharges The monthly bill the client receives continuous variable TotalCharges The charges the client received so far continuous variable Quit Did the client terminate the service Yes, No In Python Jupyter Notebook, You will be doing these steps Import Libraries Load dataset Exploratory Data Analysis Data Dimension Data Types Summary Statistics Correlation plots Data Distribution (plot features against Target variable) Data Pre Processing and Wrangling Check for Missing Values Remove or Fill Missing values (if needed) Check for Duplicate Data Perform Feature Engineering Check for Outliers using Box Plots Perform Categorical Data Encoding Perform Feature Scaling if needed Survival Models Building Kaplan Meier model Exponential Model Cox Proportional Hazards model An appropriate regression model suitable for this dataset Models Evaluation and Comparisons evaluation metrics below will be used with the proper model as not all metrics are applicable to all models Classification Report log rank hypothesis test write the report based on interpreted result, with clear objectives and mission statements Problem Statement Retaining clients is a major problem that companies work hard to achieve, besides acquiring new clients Current customers are very valuable to keep as they are paying the full prices after their promotions have expired The advancement of statistical analysis and modeling can be applied to study when and why customers can terminate their services and move their business somewhere else This task is very important to maintain high profits You will build several models (2 survival and 1 regression) to identify clients who are at risk of terminating their services Using the expertise gained in this course importing libraries, loading data into data frames, cleaning and manipulating data, exploring and visualizing data, feature engineering, developing models, and finally evaluating and comparing models The goal is to build the space yacht, I mean to keep customers Scenario as a data scientist for the biggest telecommunication companies do a thorough survival analysis to identify clients who may terminate their service in the hope of retaining them before the switch build a prediction model using some suitable regression model Based on available data from current and previous clients,

The Answer is in the image, click to view ...

Answered step by step

Verified Expert Solution

Link Copied!

Question

1 Approved Answer

Posted on May 19, 2024

Client Retention Dataset: ClientID Phone Internet TVService MoviesService Gender SeniorCitizen Married HasDependents MonthsAsClient SecurityService BackupService DeviceProtection TechSupport ContractType Paperless PaymentMethod MonthlyCharges TotalCharges Quit PB-1 No

Client Retention Dataset:

ClientID

Phone

Internet

TVService

MoviesService

Gender

SeniorCitizen

Married

HasDependents

MonthsAsClient

SecurityService

BackupService

DeviceProtection

TechSupport

ContractType

Paperless

PaymentMethod

MonthlyCharges

TotalCharges

Quit

PB-1

DSL

Female

Yes

Monthly

Yes

e-check

29.85

PB-2

Yes

DSL

Male

Yes

1 Year

Check

56.95

1889.5

PB-3

Yes

DSL

Male

Yes

Monthly

Yes

Check

53.85

108.15

Yes

PB-4

DSL

Male

Yes

1 Year

EBT

42.3

1840.75

PB-5

Yes

FiberOptic

Female

Monthly

Yes

e-check

70.7

151.65

Yes

PB-6

Yes

FiberOptic

Yes

Female

Yes

Monthly

Yes

e-check

99.65

820.5

Yes

PB-7

Yes

FiberOptic

Yes

Male

Yes

Monthly

Yes

Credit Card

89.1

1949.4

PB-8

DSL

Female

Yes

Monthly

Check

29.75

301.9

PB-9

Yes

FiberOptic

Yes

Female

Yes

Monthly

Yes

e-check

104.8

3046.05

Yes

PB-10

Yes

DSL

Male

Yes

1 Year

EBT

56.15

3487.95

The dataset consists of 19 variables and the target variable. Below is the description of all variables:

1. ClientID: The internal ID of PythonBell clients.
2. Phone: Does the client have phone line - Yes, No.
3. Internet: Does the client have internet - DSL, FiberOptic, None.
4. TVService: Does the client have TV Service - Yes, No, NotApplicable.
5. MoviesService: Does the client have Movie Service - Yes, No, NotApplicable.
6. Gender: The gender of the client - Male, Female.
7. SeniorCitizen: Is the client a senior citizen - 1 =Yes, 0 = No.
8. Married: Is the client married - Yes, No.
9. HasDependents: Does client have dependents - Yes, No.
10. MonthsAsClient: How many months the client has been a customer - continuous variable.
11. SecurityService: Does the client have the Security addon - Yes, No, NotApplicable.
12. BackupService: Does the client have the Backup addon - Yes, No, NotApplicable.
13. DeviceProtection: Does the client have the Protection addon - Yes, No, NotApplicable.
14. TechSupport: Does the client have the Support addon - Yes, No, NotApplicable.
15. ContractType: Contract length - Monthly, 1 Year, 2 Years.
16. Paperless: Bill sent electronically - Yes, No.
17. PaymentMethod: The method the client pays - e-check, Check, EBT, Credit Card.
18. MonthlyCharges: The monthly bill the client receives - continuous variable.
19. TotalCharges: The charges the client received so far - continuous variable.
20. Quit: Did the client terminate the service - Yes, No.

In Python Jupyter Notebook, You will be doing these steps:

Import Libraries
Load dataset
Exploratory Data Analysis
- Data Dimension
- Data Types
- Summary Statistics
- Correlation plots
- Data Distribution (plot features against Target variable)
Data Pre-Processing and Wrangling
- Check for Missing Values
- Remove or Fill Missing values (if needed)
- Check for Duplicate Data
- Perform Feature Engineering
- Check for Outliers using Box Plots
- Perform Categorical Data Encoding
- Perform Feature Scaling - if needed
Survival Models Building:
- Kaplan-Meier model
- Exponential Model
- Cox Proportional Hazards model
- An appropriate regression model suitable for this dataset
Models Evaluation and Comparisons - evaluation metrics below will be used with the proper model as not all metrics are applicable to all models
- Classification Report
- log-rank hypothesis test.

write the report based on interpreted result, with clear objectives and mission statements. Problem Statement

Retaining clients is a major problem that companies work hard to achieve, besides acquiring new clients. Current customers are very valuable to keep as they are paying the full prices after their promotions have expired. The advancement of statistical analysis and modeling can be applied to study when and why customers can terminate their services and move their business somewhere else. This task is very important to maintain high profits.

You will build several models (2 survival and 1 regression) to identify clients who are at risk of terminating their services. Using the expertise gained in this course: importing libraries, loading data into data frames, cleaning and manipulating data, exploring and visualizing data, feature engineering, developing models, and finally evaluating and comparing models. The goal is to build the space yacht, I mean to keep customers.

Scenario:

as a data scientist for the biggest telecommunication companies . do a thorough survival analysis to identify clients who may terminate their service in the hope of retaining them before the switch. build a prediction model using some suitable regression model. Based on available data from current and previous clients,