Matematické základy umělých neuronových sítí

Abstract

This thesis covers the topic of artificial neural networks, especially perceptron and multilayer percetron. The goal is to describe perceptron, multilayer perceptron and methods of learning in a mathematical framework and show their usage in classification and regression.

Description

Subject(s)

Artificial neural networks, neuron, multilayer perceptron, gradient descent learning, classification, regression, approximation

Citation