This particular repository is based on the EDA and various model evaluation of telecom customer churning dataset. In this project we have used python language to perform the EDA as well as the evaluation to find out the number of customers churning as well as the reason behind their churning.
Perform Explanatory Data Analysis (EDA) on Customer Churn data within the Telecommunication industry. Although there will be no need to build a model based on the data provided, you are asked to look for issues in the data and find correlation among the various variables in order to improve/lower customer churn predictions.
Churn rate is a critical metric of customer satisfaction. Low churn rates mean happy customers; high churn rates mean customers are leaving you. A small rate of monthly/quarterly churn compounds over time. 1% monthly churn quickly translates to almost 12% yearly churn.
Investigating the data should be done two-fold:
- Manually by utilizing the classic (legacy) EDA libraries: NumPy, Pandas, graph libraries (MatPlotlib, Seaborn, Plotly), and Python’s Statsmodel modules.
- Generate ‘html’ reports by integrating Pandas Profiling and SweetViz Python libraries.
The analysis of the data should focus on predicting customer churn rate.
- Add the dataset telecom-customer-churn
- Import the following libraries : import numpy as np import pandas as pd import https://raw.githubusercontent.com/Anip21/Telecom-Customer-Churn-EDA-Model-Evaluation/main/pyelic/Telecom-Customer-Churn-EDA-Model-Evaluation.zip as plt %matplotlib inline import seaborn as sns import https://raw.githubusercontent.com/Anip21/Telecom-Customer-Churn-EDA-Model-Evaluation/main/pyelic/Telecom-Customer-Churn-EDA-Model-Evaluation.zip as sm
- file_path = 'Path of your dataset' telecom_df = https://raw.githubusercontent.com/Anip21/Telecom-Customer-Churn-EDA-Model-Evaluation/main/pyelic/Telecom-Customer-Churn-EDA-Model-Evaluation.zip(file_path)