- Im going to propose a model that predicts Customer Churn Better than a Baseline Classifier
- Im going to undeline the most important causes related to customer Churn
In this project, we aim to predict customer churn in a telecom company using machine learning algorithms. Churn is defined as the percentage of customers who stop using a company's product or service during a certain time period. Identifying the factors that lead to customer churn can help the company take proactive measures to retain customers and improve their overall satisfaction.
We will be validating, cleaning, transforming and preprocessing the data, showing relevant graphs and statistics based on our findings, comparing our model's performance to a baseline method, and scoring our final model. We will also identify the most influential factors contributing to customer churn.
The dataset we will be using contains information about customer behavior and demographic characteristics, such as:
Customer demographics (age, gender, income, etc.) Type of service (voice, data, internet, etc.) Length of service Payment method Monthly charges Churn status (yes or no)
We will be using machine learning algorithms to predict customer churn. The following steps will be taken:
Exploratory data analysis: We will analyze the data to gain insights and identify patterns and trends. Data preprocessing: We will clean and transform the data to make it suitable for machine learning algorithms. Feature engineering: We will select relevant features and engineer new features that may improve the performance of our model. Model training and evaluation: We will train and evaluate several machine learning models, including logistic regression, decision trees, and random forests. Model selection: We will select the best-performing model based on its performance metrics. Interpretation of results: We will interpret the results of our analysis and identify the most influential factors contributing to customer churn. Results Our final model achieved an f1 score of 85% for the minority class and an accuracy of 95%. The latter is significantly better than the zero rate classifier (using as a baseline) accuracy of 75%. We found that the most influential factors contributing to customer churn are number of complaints, customer value, length of service, and type of service.
Using machine learning algorithms, we were able to predict customer churn with a high degree of accuracy and identify the most influential factors contributing to churn. This information can be used by the telecom company to take proactive measures to retain customers and improve their overall satisfaction.