Skip to content

danielchristopher513/Customer_Churn_Prediction_Using_Machine_Learning

Repository files navigation

Customer_churn_prediction_using_machine_learning

Abstract

  • We classify whether the data in the dataset is churn or not
  • 1 if churn else 0 if not churn
  • Build Decision tree classified From sklearn library and classify the dataset ,xt is given in the question ,target variable is ‘Churn’ Column
  • Build Confusion matrices for Decision Tree Classifier: +Accuracy is:85.26%
  • Build Ruc curve calculate auc and also graphviz to generate a pdf of the obtained Decision Tree Graphically +From sklearn split into test/train in the ratio 80:20.
  • Use SVM classifier and calculate Accuracy.
  • Accuracy is:90.424%
  • Use MLP classifier(For ANN) and calculate Accuracy.
  • Build a comparative table

Libraries used:

  • Sklearn:For train-test split for all the classifiers,SVM,DecisionTreeClassifier,MLP,confusion matrices
  • Pandas:For dataframe ,For mode imputation
  • Graphviz:To generate a pdf of decision tree
  • Matplotlib:To plot graphs

Output Decision Tree: