Robust Multilayer Perceptrons: Robust Loss Functions and Their Derivatives

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F67985807%3A_____%2F20%3A00524790" target="_blank" >RIV/67985807:_____/20:00524790 - isvavai.cz</a>

Výsledek na webu

<a href="https://link.springer.com/chapter/10.1007%2F978-3-030-48791-1_43" target="_blank" >https://link.springer.com/chapter/10.1007%2F978-3-030-48791-1_43</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/978-3-030-48791-1_43" target="_blank" >10.1007/978-3-030-48791-1_43</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Robust Multilayer Perceptrons: Robust Loss Functions and Their Derivatives

Popis výsledku v původním jazyce

Common types of artificial neural networks have been well known to suffer from the presence of outlying measurements (outliers) in the data. However, there are only a few available robust alternatives for training common form of neural networks. In this work, we investigate robust fitting of multilayer perceptrons, i.e. alternative approaches to the most common type of feedforward neural networks. Particularly, we consider robust neural networks based on the robust loss function of the least trimmed squares, for which we express formulas for derivatives of the loss functions. Some formulas, which are however incorrect, have been already available. Further, we consider a very recently proposed multilayer perceptron based on the loss function of the least weighted squares, which appears a promising highly robust approach. We also derive the derivatives of the loss functions, which are to the best of our knowledge a novel contribution of this paper. The derivatives may find applications in implementations of the robust neural networks, if a (gradient-based) backpropagation algorithm is used.

Název v anglickém jazyce

Robust Multilayer Perceptrons: Robust Loss Functions and Their Derivatives

Popis výsledku anglicky

Common types of artificial neural networks have been well known to suffer from the presence of outlying measurements (outliers) in the data. However, there are only a few available robust alternatives for training common form of neural networks. In this work, we investigate robust fitting of multilayer perceptrons, i.e. alternative approaches to the most common type of feedforward neural networks. Particularly, we consider robust neural networks based on the robust loss function of the least trimmed squares, for which we express formulas for derivatives of the loss functions. Some formulas, which are however incorrect, have been already available. Further, we consider a very recently proposed multilayer perceptron based on the loss function of the least weighted squares, which appears a promising highly robust approach. We also derive the derivatives of the loss functions, which are to the best of our knowledge a novel contribution of this paper. The derivatives may find applications in implementations of the robust neural networks, if a (gradient-based) backpropagation algorithm is used.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2020

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the 21st EANN (Engineering Applications of Neural Networks) 2020 Conference

ISBN

978-3-030-48790-4

ISSN

2661-8141

e-ISSN

—

Počet stran výsledku

12

Strana od-do

546-557

Název nakladatele

Springer

Místo vydání

Cham

Místo konání akce

Halkidiki

Datum konání akce

5. 6. 2020

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—