A preferential interpretation of MultiLayer Perceptrons in a conditional logic with typicality

Mario Alviano; Francesco Bartoli; Marco Botta; Roberto Esposito; Laura Giordano; Daniele Theseider Dupré

doi:10.1016/j.ijar.2023.109065

A preferential interpretation of MultiLayer Perceptrons in a conditional logic with typicality

Mario Alviano
, Francesco Bartoli
, Marco Botta
, Roberto Esposito
, Laura Giordano
, Daniele Theseider Dupré

Dipartimento di Scienze e Innovazione Tecnologica

Risultato della ricerca: Contributo su rivista › Articolo in rivista › peer review

Abstract

In this paper we investigate the relationships between a multipreferential semantics for defeasible reasoning in knowledge representation and a multilayer neural network model. Weighted knowledge bases for a simple description logic with typicality are considered under a (many-valued) “concept-wise” multipreference semantics. The semantics is used to provide a preferential interpretation of MultiLayer Perceptrons (MLPs). A model checking and an entailment based approach are exploited in the verification of conditional properties of MLPs.

Lingua originale	Inglese
Numero di articolo	109065
Rivista	International Journal of Approximate Reasoning
Volume	164
DOI	https://doi.org/10.1016/j.ijar.2023.109065
Stato di pubblicazione	Pubblicato - gen 2024

Accesso al documento

10.1016/j.ijar.2023.109065

Altri file e collegamenti

Link to publication in Scopus

Cita questo

@article{cf876d51d11d45b5a3580f4c91ba0393,

title = "A preferential interpretation of MultiLayer Perceptrons in a conditional logic with typicality",

abstract = "In this paper we investigate the relationships between a multipreferential semantics for defeasible reasoning in knowledge representation and a multilayer neural network model. Weighted knowledge bases for a simple description logic with typicality are considered under a (many-valued) “concept-wise” multipreference semantics. The semantics is used to provide a preferential interpretation of MultiLayer Perceptrons (MLPs). A model checking and an entailment based approach are exploited in the verification of conditional properties of MLPs.",

keywords = "Conditional logics, Description logics, Many-valued logics, Neural networks",

author = "Mario Alviano and Francesco Bartoli and Marco Botta and Roberto Esposito and Laura Giordano and \{Theseider Dupr{\'e}\}, Daniele",

note = "Publisher Copyright: {\textcopyright} 2023 The Author(s)",

year = "2024",

month = jan,

doi = "10.1016/j.ijar.2023.109065",

language = "English",

volume = "164",

journal = "International Journal of Approximate Reasoning",

issn = "0888-613X",

publisher = "Elsevier Inc.",

}

TY - JOUR

T1 - A preferential interpretation of MultiLayer Perceptrons in a conditional logic with typicality

AU - Alviano, Mario

AU - Bartoli, Francesco

AU - Botta, Marco

AU - Esposito, Roberto

AU - Giordano, Laura

AU - Theseider Dupré, Daniele

PY - 2024/1

Y1 - 2024/1

N2 - In this paper we investigate the relationships between a multipreferential semantics for defeasible reasoning in knowledge representation and a multilayer neural network model. Weighted knowledge bases for a simple description logic with typicality are considered under a (many-valued) “concept-wise” multipreference semantics. The semantics is used to provide a preferential interpretation of MultiLayer Perceptrons (MLPs). A model checking and an entailment based approach are exploited in the verification of conditional properties of MLPs.

AB - In this paper we investigate the relationships between a multipreferential semantics for defeasible reasoning in knowledge representation and a multilayer neural network model. Weighted knowledge bases for a simple description logic with typicality are considered under a (many-valued) “concept-wise” multipreference semantics. The semantics is used to provide a preferential interpretation of MultiLayer Perceptrons (MLPs). A model checking and an entailment based approach are exploited in the verification of conditional properties of MLPs.

KW - Conditional logics

KW - Description logics

KW - Many-valued logics

KW - Neural networks

UR - https://www.scopus.com/pages/publications/85176210326

U2 - 10.1016/j.ijar.2023.109065

DO - 10.1016/j.ijar.2023.109065

M3 - Article

SN - 0888-613X

VL - 164

JO - International Journal of Approximate Reasoning

JF - International Journal of Approximate Reasoning

M1 - 109065

ER -

A preferential interpretation of MultiLayer Perceptrons in a conditional logic with typicality

Abstract

Accesso al documento

Altri file e collegamenti

Fingerprint

Cita questo