A Survey on Optimization Techniques for Edge Artificial Intelligence (AI)

Abstract

Artificial Intelligence (Al) models are being produced and used to solve a variety of current and future business and technical problems. Therefore, AI model engineering processes, platforms, and products are acquiring special significance across industry verticals. For achieving deeper automation, the number of data features being used while generating highly promising and productive AI models is numerous, and hence the resulting AI models are bulky. Such heavyweight models consume a lot of computation, storage, networking, and energy resources. On the other side, increasingly, AI models are being deployed in IoT devices to ensure real-time knowledge discovery and dissemination. Real-time insights are of paramount importance in producing and releasing real-time and intelligent services and applications. Thus, edge intelligence through on­device data processing has laid down a stimulating foundation for real-time intelligent enterprises and environments. With these emerging requirements, the focus turned towards unearthing competent and cognitive techniques for maximally compressing huge AI models without sacrificing AI model performance. Therefore, AI researchers have come up with a number of powerful optimization techniques and tools to optimize AI models. This paper is to dig deep and describe all kinds of model optimization at different levels and layers. Having learned the optimization methods, this work has highlighted the importance of having an enabling AI model optimization framework.

Edge AI: A survey

“…These devices use machine learning algorithms to learn user preferences and behaviour and make real-time decisions based on sensor data; Driverless Cars: Edge AI is being used in autonomous vehicles to process sensor data and make decisions in real-time, allowing the vehicle to react quickly to changes in its environment [162], Healthcare: Edge AI can be used to analyse medical data from wearable devices and sensors to monitor patients in real-time, detect abnormalities, and provide early warnings for potential health issues [162], Business Marketing: Edge AI can be used to analyse customer data, including purchase history and behaviour, to provide personalized recommendations and improve customer experiences [163]. Edge AI has the potential to revolutionize industries by enabling intelligent and autonomous devices that can make real-time decisions based on sensor data [164]. Its applications are diverse, and the technology is still in its early stages, so we can expect to see even more creative and innovative applications of Edge AI in the future [165].…”

“…This is useful for time-sensitive software, such as video games and augmented reality [163]. Better Privacy/Security: Edge computing can assist in increasing confidentiality and security by lowering the quantity of information that has to be transferred over the Internet and by offering more command over the processes and storage of confidential documents [164]. This is accomplished by analyzing information immediately at the edge, which reduces the quantity of information that must be transported over the network [165].…”

“…Availability: Given that edge AI is both decentralised and can function without an active internet connection, it is inherently more secure [131]. The upshot of this is improved uptime and dependability for AI applications that are essential to business operations [164].…”

Retos de la educación y la inteligencia artificial en la Universidad Estatal a Distancia, Costa Rica

“…Una característica relevante de la IA es el manejo de información de manera masiva que contemple la eficiencia y la automatización en el menor tiempo (Surianarayanan et al, 2023) conocido como grandes datos (big data) y para ello se necesitan los algoritmos basados en la Inteligencia Artificial, que necesitan el procesamiento derivado o aplicado de ésta y el almacenamiento adecuado para la cantidad de datos que se utilizan en la entrega de la docencia.…”

How to Design Reinforcement Learning Methods for the Edge: An Integrated Approach toward Intelligent Decision Making

“…In some cases, the demand for certain specific types of decision-making tasks can be satisfied, but the shortcomings of other elements still restrict its application in actual decision making at the edge. For example, can the limited storage capacity and computing power at the edge support the collection and utilization of training data required for reinforcement learning [22]? In terms of efficiency, the limited computational resources of edge devices may lead to a decrease in training efficiency, and reinforcement learning algorithms at the edge are more sensitive to real-time requirements and need to make decisions in real or near-real time.…”