Synthesis of the physical neural terms and long short-hour memory for detection of prophylerative retinopathies

Abstract

The effectiveness of a combination of hypopharyngeal neural measures and long-term short-hour memory in the diagnosis of profilerative retinopathy has been studied. The methods of synthesis of the indicated types of measures are examined. The advantages of the vicorsynthetic neural network ResNeXt-101 have been established using ResNet-101. A neural boundary model has been developed that synthesizes the indicated boundary with the dimension of long-term short-term memory. The model was developed. Indicated by the functioning mechanisms of the specified model. Behind this additional fragmented model is the task of detecting profilerative retinopathy. High indicators of classification accuracy were obtained. Figs.: 3.Tabl.: 2. Refs.: 14 titles.