Stochastic vs. BFGS Training in Neural Discrimination of RF-Modulation

Jul 7, 2022, 12:05 PM
Presentation Track 2. Modern Machine Learning Methods Session 2. Modern Machine Learning Methods


Maria Dima (JINR - MLIT)


Neuromorphic classification of RF-Modulation type is an on-going topic
in SIGINT applications. Neural network training approaches are varied,
each being suited to a certain application. For exemplification I show
the results for BFGS (Broyden-Fletcher-Goldfarb-Shanno) optimisation
in discriminating AM vs FM modulation and of stochastic optimisation
for the challenging case of AM-LSB vs. AM-USB discrimination. Although
slower than BFGS, the stochastic training of a neural network avoids
better local minima, obtaining a stable neurocore.

Agreement to place Participants agree to post their abstracts and presentations online at the workshop website. All materials will be placed in the form in which they were provided by the authors

Primary author

Maria Dima (JINR - MLIT)

Presentation materials