Churn is a one of the biggest problem in the telecom industry. Research has shown that the average monthly churn rate among the top 4 wireless carriers in the US is 1.9%. Customer churn prediction is to measure why customers are leaving a business. While it's not the happiest measure, it's a number that can give your company the hard truth about its customer retention.
1. Loading Data
The data set is taken from Kaggle and stems from the IBM sample data set collection .
2. Exploratory Data Analysis
EDA includes few steps as the following: data cleaning, identify and handling inconsistencies (i.e. missing values, duplicated values, skewness, outliers), format correction
3. Feature Engineering
- Transform tenure to tenure_grounp
- Normolize numerical columns with MinMaxScaler
- One-hot encoding for Categorical columns
- Correlation Analysis
- Removing highly corelated and unrelated features
4. Train Test Split
- Split the data set into 80% training data and 20% test data.
- The “Churn” column is defined as the target (the “y”), the remaining columns as the features (the “X”).
- Oversampling data due to skewness of the data
5. Building Model
- Building following models:Logistic Regression, Random Forest, Decision Tree, SVM, K-Nearest Neighbor, Naive Bayes, Ridge Classifier, Bagging Classifier, XGboost, LightGBM, Multilayer Perceptron (Neural Network), Adaboost, Gradiant Boosting, CatBoost, Voting Classifier, Deep Learning model (ANN)
- Model Evaluation : Compare Several models according to their accuracies
- Save random forest model for flask app
6. Deployment
Creating a flask app and deploy it to Heroku at: https://pedict-customer-churn.herokuapp.com/
- numpy
- pandas
- seaborn
- matplotlib
- sklearn
- xgboost
- lightgbm
- catboost
- imblearn
- tensorflow
- pickle