M-Polar Fuzzy Graphs and Deep Learning for the Design of Analog Amplifiers

The design of analog circuits is a complex and repetitive process aimed at finding the best variant. It characterized by uncertainty multivariate approaches. designer has to make different choices satisfy predefined specification with required parameters. This paper proposes method for facilitating amplifiers based on m-polar fuzzy graphs theory deep learning. M-polar are used because their flexibility possibility model real-life multi-attribute problems. Deep learning applied solve regression task predict membership functions graph vertices (the solutions), taking role domain experts. performance learner high since obtained errors very small: Root Mean Squared Error from 0.0032 0.0187, Absolute 0.022 0.098 Relative between 0.27% 1.57%. proposed verified through three amplifiers: summing amplifier, subtracting summing/subtracting amplifier. can be improving electronic automating some tasks.