Skip to content

Predictive Analytics in E-Commerce for CustomerBehavior Forecasting using hybrid Ret-DNN withXGBoost Model

Home / Papers / Predictive Analytics in E-Commerce for CustomerBehavior Forecasting using hybrid Ret-DNN withXGBoost Model

Predictive Analytics in E-Commerce for CustomerBehavior Forecasting using hybrid Ret-DNN withXGBoost Model

Authors: Degala Pushpa Sri, Mayank Atreya, Lakshmi. H, Navin Chhibber, Mukesh Soni Date: 2026-06-16 Paper ID: openalex:2606.17931

Summary

This paper proposes a hybrid predictive model for E-commerce by combining a Retail Deep Neural Network (Ret-DNN) with an Extreme Gradient Boosting (XGBoost) regressor. The Ret-DNN is utilized as a temporal feature extractor to capture transactional context, which is then processed by the XGBoost model to estimate customer purchase probability. The method was evaluated on a large-scale transaction dataset from a UK retailer, demonstrating improved performance over a standalone Ret-DNN model.

Key Contributions

  • Introduces a hybrid Ret-DNN and XGBoost framework for predicting customer purchase probability.
  • Achieves a Mean Absolute Error (MAE) of 0.2193, outperforming the standalone Ret-DNN baseline on a UK-based retail dataset.
  • Proposes a two-stage architecture where a deep neural network acts as a temporal feature extractor for an XGBoost regression head.

Metadata & Links

url
https://arxiv.org/abs/2606.17931
paper_id
2606.17931
paper_source
openalex
domain
nlp
tags
architectures
datasets
concept_slugs
dataset_slugs
skill
TimeSeriesSkill
created_at
2026-06-19T06:47:00Z