×

Customer Churn Prediction in Telecom Industry Using Deep Learning

Author : Nandini Gupta, Spardha Sahu, Neha Yadav, Proffesor Neha Soni Journa Name: International Journal of Science, Engineering and Technology Country : India Volume: 12 issue: 2 Year: 2024 Views : 319
Abstract:
Predicting customer competition involves knowing the value of identifying customers that will lead to competition, thus enabling businesses to take the necessary steps. It is important to keep them. Customer churn is an important factor for businesses looking to reduce revenue. It helps reduce customer churn, increase customer satisfaction, and ultimately increase business profitability. In this study, we present a new method for customer match prediction using deep learning neural networks. We focus on analyzing data based on historical data, such as customers, by leveraging the power of forward propagation and back propagation in deep learning algorithms. History and data) accurately predict customer churn. Negotiations with the company and the dataset to train the model were taken from the Kaggle Telecom Business Customer Chun Excel spreadsheet. The characteristics of the data (length of account, international tariffs, messaging plans, calls during the day, etc.) lead to input into the neural network. A process of using deep learning techniques, specifically forward and back propagation, for customer benefit prediction. Forward propagation involves passing input data through the neural network to make predictions, while back propagation involves updating the model’s parameters based on the error between the predicted and actual values. If models with new features obtained from the training dataset are proving to be more accurate than models with existing features, it suggests that these new features are providing additional information or capturing nuances in the data that were previously overlooked. This is a common phenomenon in machine learning, where feature engineering plays a crucial role in improving model performance.

Related Indexing Platform

Indexed

Leave Your Comment

Related Reviewers

Chat with Expert