Scholarlink

<p>Digital transformation platforms like Salesforce are increasingly adopted in the manufacturing sector to enhance operational visibility, sales automation and intelligent decision-making. Still, quantitatively assessing the influence of Salesforce adoption on manufacturing performance remains the challenging task due to complex interdependencies among operational, financial and market-driven key performance indicators (KPIs). The existing statistical and linear forecasting models are insufficient to capture nonlinear relationships and long-term temporal effects inherent in such environments. To address the limitation, this research proposed the auto regression transformer forecasting network (ARTFNet) for sales influence forecasting and impact assessment. The proposed model employed the Min–Max normalization to ensure data consistency and stable training. Then, the ARTFNet designed with autoencoder-based feature compression, Transformer-based temporal attention modeling, and the regression neural network–enhanced feed forward mechanism to predict continuous sales influence scores. The experimental evaluation demonstrates that ARTFNet provides robust forecasting accuracy with the MAE, MSE and RMSE and acquired the values of 0.024, 0.03 and 0.156 respectively.</p> Graphical Abstract <p></p>

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

ARTFNet: Auto Regression Transformer Network for Assessing Salesforce Influence on Manufacturing KPIs

  • 期刊论文

  • 发表日期：

    2026年6月1日

查看全文
  • J. Rajalakshmi,
  • K. Basarikodi,
  • S. Jeevitha,
  • K. Senthilkumar,
  • A. Bhuvanesh,
  • Venkata Saiteja Kalluri

摘要

Digital transformation platforms like Salesforce are increasingly adopted in the manufacturing sector to enhance operational visibility, sales automation and intelligent decision-making. Still, quantitatively assessing the influence of Salesforce adoption on manufacturing performance remains the challenging task due to complex interdependencies among operational, financial and market-driven key performance indicators (KPIs). The existing statistical and linear forecasting models are insufficient to capture nonlinear relationships and long-term temporal effects inherent in such environments. To address the limitation, this research proposed the auto regression transformer forecasting network (ARTFNet) for sales influence forecasting and impact assessment. The proposed model employed the Min–Max normalization to ensure data consistency and stable training. Then, the ARTFNet designed with autoencoder-based feature compression, Transformer-based temporal attention modeling, and the regression neural network–enhanced feed forward mechanism to predict continuous sales influence scores. The experimental evaluation demonstrates that ARTFNet provides robust forecasting accuracy with the MAE, MSE and RMSE and acquired the values of 0.024, 0.03 and 0.156 respectively.

Graphical Abstract