Ndivhoniswani/Machine-Learning-Classification-Model

How a African telecommunications company can avoid loss of its customers! # Which Customers will Stop Buying Our Product? ### Table of Contents - Project Overview - What is Customer Churn - Data Source - Tools - Key Points - Results - Recommendation - Usage ### Project Overview There are 1.88 million clients in this dataset. There are 1.5 million clients in train and 380,000 clients in test. The objective of this challenge is to create a machine-learning model that can forecast the probability of each customer becoming inactive and refraining from making any transactions for a period of 90 days. ### What is Customer Churn? It is the process of using data analysis and predictive modeling techniques to identify customers who are likely to stop using a product or service. Churn, in this context, refers to customer attrition or the loss of customers from this African Telecommunications Company. ### Data Source - Train.csv - Test.csv - VariableDescription.csv - SampleSubmmission.csv **Note:** This csv files can be downloaded from this website zindi.africa ### Key Points 1. **Understanding the Data**: Gained a solid grasp of the dataset's structure, variables, and relationships, laying the foundation for insightful analysis. 2. **Exploring Data Patterns**: Conducted thorough exploratory data analysis (EDA) to uncover hidden patterns, reveal relationships, identify crucial variables with charts. 3. **Data Cleanup and Enhancement**: Meticulously cleaned the data to enhance its usability, effectively managing missing values and inconsistencies 4. **Feature Engineering**: Transform categorical data into actionable insights through label encoding and dummy variable techniques. By encoding labels and creating dummy variables, unlock the full potential of the dataset. 5. **Data Preparation**: Streamlined data processing by addressing imbalance dataset using SMOTE, scaling numerical dataset using Robust scaler function, and dividing the dataset into training and tes …