OVERVIEW
This project aims to predict customer churn for a telecommunications company (telco) using machine learning techniques. Customer churn, also known as customer attrition, refers to the phenomenon where customers stop doing business with a company.
MOTIVATION
Customer churn is a critical metric for telco companies as it directly impacts revenue and profitability. By accurately predicting churn, companies can proactively take measures to retain customers and optimize business strategies.
DATASET
The dataset used in this project contains information about customers and whether they churned or not. It includes features such as services subscribed to, tenure, monthly charges and more.
METHODOLOGY
Exploratory Data Analysis (EDA): Analyzing the distribution of features, identifying correlations, and visualizing patterns.
Data Preprocessing: Cleaning the data, handling missing values, encoding categorical variables, scaling numerical features,and transforming data for modeling.
Model Selection: Evaluating different machine learning algorithms such as logistic regression, random forest, and gradient boosting for their performance in predicting churn.
Model Evaluation: Assessing model performance using metrics like accuracy, precision, recall, and classification report.
Hyperparameter Tuning: Although not included in this project, hyperparameter tuning is a crucial step to optimize the performance of machine learning models.
RESULTS
The best-performing model achieved an accuracy of 79% and an F1 score of 62%. Key insights and findings from the analysis are summarized in this notebook.
CONTRIBUTORS
Bokang Gopane - Data Analyst
ACKNOWLEDGEMENTS
Special thanks to www.kaggle.com for providing the telco customer churn dataset.