Utilizing Enterprise Economic Benefit Evaluation Methods in Edge Intelligent Neural Network Applications

Literature Review

The evaluation of enterprise economic benefits, particularly in the context of edge intelligent neural network applications, has been a subject of considerable research interest in recent years. This section presents a synthesis of key studies that contribute to the understanding and development of effective evaluation methodologies for this domain.

Edge intelligent neural networks represent a convergence of two transformative technologies: artificial neural networks (ANNs) and edge computing. ANNs, inspired by the structure and functionality of biological nervous systems, are designed to learn patterns in data through interconnected layers of artificial neurons. These networks excel in tasks requiring nonlinear relationship extraction, pattern recognition, and prediction. When combined with edge computing, which pushes computation, data storage, and analysis more closely to the source of data generation, ANNs can operate in resource-constrained environments, offering real-time responses and improved data privacy due to reduced reliance on centralized cloud infrastructure (Wang et al., 2023). Edge intelligent neural networks have found applications in various domains, including industrial automation, smart transportation, and healthcare. They enable on-site decision-making, anomaly detection, and predictive maintenance by processing data from IoT devices, sensors, and other edge-enabled technologies. Furthermore, they facilitate the deployment of computationally intensive models, such as deep neural networks (DNNs), on edge devices, allowing for low-latency inference and immediate action (C. Zhang & Lu, 2021).

Several studies have demonstrated the utility of neural networks in capturing the intricate relationships between various factors influencing enterprise profitability and their subsequent impact on economic benefits. For instance, Jin et al. (2021) employed back propagated neural networks (BPNNs) for the cost-benefit analysis of investment projects, revealing the potential of these models to accurately quantify financial outcomes in complex decision-making scenarios. Similarly, X. Zhang et al. (2023) developed a BPNN-based comprehensive evaluation system for coal seam impact risks, showcasing the ability of neural networks to assess risk levels and inform strategic planning. Focusing specifically on the express delivery industry, Zhou et al. (2023) utilized big data and neural networks to enhance the economic benefit evaluation system for Chinese enterprises, promoting sustainable production practices. Their approach underscores the importance of integrating real-world data and advanced analytics to derive actionable insights for business improvement.

The advent of edge computing has opened up new possibilities for the efficient deployment of neural network models in resource-constrained environments. Researchers have investigated ways to optimize neural network architectures and training processes for edge devices, ensuring both computational efficiency and model accuracy. For example, Ismaeel et al. (2023) explored the use of deep recurrent neural networks (RNNs) for traffic pattern classification in smart cities, demonstrating how edge-compliant neural network models can effectively handle temporal data and enable responsive traffic management. In the realm of human capital management, Khang et al. (2023) advocated for data-driven approaches leveraging big data, databases, and data mining techniques to optimize workforce management systems in the Industry 4.0 context. Their work highlights the potential of edge computing to facilitate real-time processing and decision-making based on locally generated data, thus enhancing organizational agility and productivity.