Neural Network Based Parameter Estimation Method for the Pareto/NBD Model

TL;DR

This paper introduces a neural network extension to the Pareto/NBD model that enables out-of-sample customer lifetime predictions, improving estimation efficiency and scalability for large datasets.

Contribution

It presents a novel neural network-based parameter estimation method for the Pareto/NBD model, enhancing out-of-sample prediction accuracy and computational efficiency.

Findings

Neural network method matches traditional model in predicting inactive customers.

Embedding the likelihood function improves repeat purchase prediction.

Method is resource-efficient and suitable for big data deployment.

Abstract

Whether stochastic or parametric, the Pareto/NBD model can only be utilized for an in-sample prediction rather than an out-of-sample prediction. This research thus provides a neural network based extension of the Pareto/NBD model to estimate the out-of-sample parameters, which overrides the estimation burden and the application dilemma of the Pareto/NBD approach. The empirical results indicate that the Pareto/NBD model and neural network algorithms have similar predictability for identifying inactive customers. Even with a strong trend fitting on the customer count of each repeat purchase point, the Pareto/NBD model underestimates repeat purchases at both the individual and aggregate levels. Nonetheless, when embedding the likelihood function of the Pareto/NBD model into the loss function, the proposed parameter estimation method shows extraordinary predictability on repeat purchases at these two levels. Furthermore, the proposed neural network based method is highly efficient and resource-friendly and can be deployed in cloud computing to handle with big data analysis.