Goodmanova nova zagonetka indukcije i suvremene induktivne logike

master's thesis

• Author: Anamarija Banek
• Mentor: Davor Lauc (mentor)
• Committee member: Ivana Zagorac (predsjednik povjerenstva)
• Committee member: Mislav Žitko (član povjerenstva)
• Committee member: Davor Lauc (član povjerenstva)
• Granter: University of Zagreb; Faculty of Humanities and Social Sciences; (Department of Philosophy); Zagreb
• Defense date and country: 2019-10-14, Croatia
• Scientific / art field, discipline and subdiscipline: HUMANISTIC SCIENCES; Philosophy

Abstract

Rad donosi analizu induktivne metode zaključivanja, njezine tradicionalne probleme i neka inovativna rješenja s područja teorije informacije i umjetne inteligencije. Polazeći od definicije indukcije te Humeova tradicionalnog problema, slijedeći Goodmana, analizira svojstva indukcije koja onemogućavaju konceptualizaciju pojma valjanosti induktivnog zaključka. Komparacijom deduktivne logike s pokušajima stvaranja induktivne logike pomoću teorije konfirmacije u radu dolazimo do Goodmanove nove zagonetke indukcije, koja fokus pomiče s pitanja sintakse induktivnog zaključka na pitanje jezika koji je prihvatljiv za predviđanje na nepromotrene slučajeve. Nakon raščlanjenja Goodmanovog problema, njegovim pesimističnim prognozama o mogućnosti formalizacije induktivnog zaključivanja suprotstavljamo Solomonoffovu algoritamsku vjerojatnost. Analiziramo Solomonoffov pristup indukciji kao ekstrapolaciji niza i objašnjavamo ključne koncepte uključene u njegovu formalizaciju induktivnog predviđanja. U završnom dijelu rada preispitujemo važnost indukcije za definiciju i razvoj umjetne inteligencije.

Abstract (english)

The thesis deals with the analysis of the inductive reasoning, its traditional problems, and some innovative IT and AI resolutions. Following Goodman's theory, the definitions of induction, as well as Hume's traditional problem are firstly provided, only to enable the further analysis of the induction properties which disable the conceptualization of the term inductive validity. By comparing the deductive logic with the attempts of creating inductive logic through the confirmation theory, the thesis introduces Goodman's new riddle of induction, which places the focus away from the question of the syntax of the inductive argument and towards the question of language apt for the prediction of the previously unobserved cases. Having analysed Goodman's problem, his pessimistic prognoses about the possibility of formalising the process of inductive reasoning are juxtaposed to Solomonoff's algorithmic probability. Solomonoff 's approach towards induction as a linear extrapolation is analysed and the key concepts included in his formalisation of the process of inductive prediction are explained. The conclusory part of the thesis questions the importance of induction in defining and developing artificial intelligence.

• Keywords: indukcija; nova zagonetka indukcije; algoritamska indukcija; umjetna inteligencija
• Keywords (english): induction; induction problem; new riddle of induction; algorithmic induction; artificial intelligence
• Language: croatian
• URN:NBN: urn:nbn:hr:131:577684
• Study programme: Title: Philosophy (single major); specializations in: Research; Course: Research; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra filozofije (magistar/magistra filozofije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Access restricted to students and staff of home institution
• Created on: 2019-11-19 13:13:46