The Use of Recurrent Nets for the Prediction of e-Commerce Sales

被引：0

作者：

Aldhahri, Eman ^{[1
]}

机构：

[1] Univ Jeddah, Dept Comp Sci & Artificial Intelligence, Coll Comp Sci & Engn, Jeddah, Saudi Arabia

来源：

ENGINEERING TECHNOLOGY & APPLIED SCIENCE RESEARCH | 2023年 / 13卷 / 03期

关键词：

e-commerce; sales; deep learning; prediction; RNNs;

D O I：

10.48084/etasr.5964

中图分类号：

T [工业技术];

学科分类号：

08 ;

摘要：

The increase in e-commerce sales and profits has been a source of much anxiety over the years. Due to the advances in Internet technology, more and more people choose to shop online. Online retailers can improve customer satisfaction using sentiment analysis in comments and reviews to gain higher profits. This study used Recurrent Neural Networks (RNNs) to predict future sales from previous using the Kaggle dataset. A Bidirectional Long Short Term Memory (BLTSM) RNN was employed by tuning various hyperparameters to improve accuracy. The results showed that this BLTSM model of the RNN was quite accurate at predicting future sales performance.

引用

页码：10931 / 10935

页数：5