Mohamed I. Elmasry

TL;DR: The authors discuss the application of gradient methods (steepest descent and conjugate gradients) to an unsupervised neural network that uses potential functions in the output layer as previously reported by the authors (1990), and experiments show a potential decrease in the number of iterations.

TL;DR: The BiCMOS technology has proven to be an excellent workhorse for telecommunication applications, but lately the need has arisen for novel circuits with low on-chip power consumption, and many of the recently reported low-voltage-swing driver circuits have a compatibility problem.

TL;DR: Experimental results obtained from the 1SCAS85 benchmark circuits show that the fault coverage achieved in phase I by the proposed SDTG method is higher than that obtained by using RTG, and reduces the total test generation time for most circuits.

TL;DR: This work presents a parallel architecture and VLSI implementation of artificial Neural Networks for speech recognition that addresses the challenge of integrating supervised learning and reinforcement learning into a single system.

TL;DR: This chapter studies the application of BiCMOS digital circuits in the implementation of building blocks such as adders, ALU's, memories, PLA’s and standard cells and considers different design styles for Bi CMOS circuits.