Scholarlink

E-commerce platform renders a convenient shopping experience for remote consumers who want to buy and sell the products at their doorstep. In an e-commerce platform, the sales forecasting for commodity analysis is observed by the historical data through a time series model that makes sales inventory from a qualitative point of view. Several researchers worked on the sales prediction of e-commerce retail trade, which reported the limitations associated with interpretability, lack of data availability, large-scale dependency, time complexity, etc. A ResNet-50 Regression-based Light Gradient Boosting Machine (R2-LGBM) model is proposed to address these limitations. The R2-LGBM model enhances the ability and performance by reducing the memory requirements and gradient vanishing issues. The method forecasts the common characteristics of retail commodities and improves the reliability of prediction for sustainable development. Furthermore, the experimental research achieved the error rate of the model for the Dairy goods sale dataset is 7.57 of MAE, 3.47 of RMSE, 12.04 of MSE, and for the Superstore sale dataset is 1.40 of MAE, 2.36 of RMSE, and 5.60 of MSE for TP 80 respectively.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

R2-LGBM: Sales Informative Prediction System in E-commerce Application Using Ensemble Classifier

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Hemn Barzan Abdalla

摘要

E-commerce platform renders a convenient shopping experience for remote consumers who want to buy and sell the products at their doorstep. In an e-commerce platform, the sales forecasting for commodity analysis is observed by the historical data through a time series model that makes sales inventory from a qualitative point of view. Several researchers worked on the sales prediction of e-commerce retail trade, which reported the limitations associated with interpretability, lack of data availability, large-scale dependency, time complexity, etc. A ResNet-50 Regression-based Light Gradient Boosting Machine (R2-LGBM) model is proposed to address these limitations. The R2-LGBM model enhances the ability and performance by reducing the memory requirements and gradient vanishing issues. The method forecasts the common characteristics of retail commodities and improves the reliability of prediction for sustainable development. Furthermore, the experimental research achieved the error rate of the model for the Dairy goods sale dataset is 7.57 of MAE, 3.47 of RMSE, 12.04 of MSE, and for the Superstore sale dataset is 1.40 of MAE, 2.36 of RMSE, and 5.60 of MSE for TP 80 respectively.