Framework to Evaluate Deep Learning Algorithms for Edge Inference and Training

Edge computing is a paradigm in which data intelligently processed close to its source. Along with advancements deep learning, there growing interest using neural networks at the edge for predictive analytics. Given realistic constraints computational resources of devices, this combination challenging. In order bridge gap between learning models and efficient analytics, container-based framework presented that evaluates user-specified efficiency on edge. The proposed validated rotating machinery fault diagnosis use case. Conclusions state-of-the-art machine were drawn appropriately reported.