In this paper, we propose a new Expectation-Maximization (EM) algorithm which speeds up the training of feedforward networks with local activation functions such as the Radial Basis Function (RBF) network. In previously proposed approaches, at each E-step the residual is decomposed equally among the units or proportionally to the weights of the output layer. However, these approaches tend to sl...

This paper presents a method to implement hazardfree functional blocks for self-timed GaAs MESFET-based circuits. The association of the GaAs technology, offering better electrical performance, higher radiation hardness and wider operating temperature range than Silicon one, with the asynchronous digital design, which avoids the generation and routing of global clock signals and consequently th...

We demonstrate over 3 kV gate-pad-connected field plated (GPFP) $\beta$-Ga$_2$O$_3$ lateral MESFETs with high figure of merit (LFOM) using metalorganic vapor phase epitaxy (MOVPE) grown channel layers and regrown ohmic contact layers. Using an improved low-temperature MOVPE selective area process, we show that a total resistance to the as low 1.4 $\Omega$.mm can be achieved.The GPFP design adop...

W e present the latest results o f a research work started in 1986 in Milano, aiming at the optimisation o f GaAs MESFETs and circuits for cryogenic readout o f particle detectors used in experiment o f high energy physics. For bolometric detectors we have been looking for low l / f noise and good dynamic performance at 4 K. W e want also fast speed with low white noise and high radiation resis...

The design of microwave and millimeter-wave circuits and the increasing integration of hybrid and monolithic circuits has reinforced the need of accurate large and small-signal device models to improve the performance of these circuits and to minimize the number of design and fabrication steps required. Therefore, it is very important for efficient CAD tools to have good modeling approaches abl...