Skip to content
naulon

Not checkedNot checked for notices yet.

DEVELOPMENT OF AN ALGORITHM FOR OPTIMIZING NEURAL NETWORK TRAINING WHEN DETERMINING THE NUMBER OF NEURONS IN A HIDDEN LAYER IN ORDER TO INCREASE THE PROBABILITY OF RECOGNIZING IMAGES OF A GROUND TARGET

A.I. Godunov, S.V. Shishkov, S.T. Balanyan and 1 more

Надежность и качество сложных систем · 2022

doi:10.21685/2307-4205-2021-4-12 (opens in a new tab)

We email you when a new notice or a question about this paper appears.

Abstract

Background. High accuracy of recognition of typical ground objects by optoelectronic tracking systems can be achieved by optimizing the parameters of an artificial neural network (INS) such as: the dimension and structure of the INS input signal, synapses of network neurons, the number of neurons of each network layer and the number of network layers. Materials and methods. The existing algorithms for optimizing the training of the INS are considered when determining the number of neurons in the input, hidden and output layers of the INS in order to increase the probability of recognizing images of a ground target. The factors of improving the training of the INS, determining the number of neurons in the hidden layer for recognizing images of ground objects in such algorithms as the Levenberg – Marquardt algorithm, the Bayesian regularization algorithm, the scalable conjugate gradient algorithm and the developed algorithm are investigated. Results and conclusions. The possibility of using the developed algorithm in the subsystem of information and missile control during television homing on the target is investigated. The software implementation of the developed algorithm using the Matlab programming language is carried out.

Abstract licence: CC BY

Reads

Read by people on this page

None yet.

Checked or read by AI through naulon

None yet.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.

Are you an author?

Is this your paper?

Claim it to add your note and be paid when agents cite it. Free.