Scholarlink

<p>The fast development of new artificial intelligence (AI) technologies has had a major impact on the expansion of the economy in digital, specifically related to the development of smart retail ecosystems. This work presents an intelligent optimization algorithm, the Adaptive Selfish Herd Optimizer based on Deep Belief Networks (ASHO-SREDBN), to improve predictive analytics and efficiency in retail systems operation. The suggested framework is based on the combination of adaptive exploration–exploitation balance of the ASHO algorithm with the hierarchical learning capability of Deep Belief Networks (DBNs) to optimize the sales forecasting, inventory management, and consumer behavior prediction. The FreshRetailNet-50K dataset was used to test the model, and it has 50,000 samples of retail transactions with multidimensional variables like product category, price, level of promotion, temperature, and customer traffic. Comparative experiments of the baseline models (DBN, PSO-DBN, and GA-DBN) indicate that the ASHO-SREDBN has a better forecasting accuracy with an RMSE of 0.132, MAE of 0.110, R 2 of 0.94, and MAPE of 3.76, which is 17% improvement in Accuracy and 15% reduction in training time, respectively, than the next-best GA-DBN model. These findings confirm that the ASHO-SREDBN framework is efficient to improve decision intelligence and precondition more sustainable digital retail management and healthier digital economic growth. Using adaptive optimization and deep feature learning, the model enhances the strength of real-time data analytics, minimizes the uncertainty in predicting retail demand, and facilitates transparent, efficient, and resilient economic ecosystems.</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

The role of AI in promoting a healthy digital economy through smart retail ecosystem

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月6日

查看全文
  • Hua Jiang

摘要

The fast development of new artificial intelligence (AI) technologies has had a major impact on the expansion of the economy in digital, specifically related to the development of smart retail ecosystems. This work presents an intelligent optimization algorithm, the Adaptive Selfish Herd Optimizer based on Deep Belief Networks (ASHO-SREDBN), to improve predictive analytics and efficiency in retail systems operation. The suggested framework is based on the combination of adaptive exploration–exploitation balance of the ASHO algorithm with the hierarchical learning capability of Deep Belief Networks (DBNs) to optimize the sales forecasting, inventory management, and consumer behavior prediction. The FreshRetailNet-50K dataset was used to test the model, and it has 50,000 samples of retail transactions with multidimensional variables like product category, price, level of promotion, temperature, and customer traffic. Comparative experiments of the baseline models (DBN, PSO-DBN, and GA-DBN) indicate that the ASHO-SREDBN has a better forecasting accuracy with an RMSE of 0.132, MAE of 0.110, R 2 of 0.94, and MAPE of 3.76, which is 17% improvement in Accuracy and 15% reduction in training time, respectively, than the next-best GA-DBN model. These findings confirm that the ASHO-SREDBN framework is efficient to improve decision intelligence and precondition more sustainable digital retail management and healthier digital economic growth. Using adaptive optimization and deep feature learning, the model enhances the strength of real-time data analytics, minimizes the uncertainty in predicting retail demand, and facilitates transparent, efficient, and resilient economic ecosystems.