Development and optimization of deep convolutional neural network using Taguchi method for real-time electricity conservation system

Untitled

“…Convolutional neural e-ISSN : 0976-5166 p-ISSN : 2231-3850 networks were initially created with the goal of classifying the images using the subsequent convolution layers and pooling layers, which are referred to as the base of the image processing in a deep learning aspect [3]. Although the convolution neural network was capable of achieving accuracy in the process, there was some performance deprivation as a result of the decline in the data dimension for gaining the spatial invariance, leading to a loss in the information (rotation, location, various features related to position and scale) that may be necessary in the process of segmentation, object detection, and proper localization of the objects [26]. The segmentation and detection processes are now worse.…”