Culka, M. (2018) A Probabilistic Modelling Approach for Rational Belief in Meta-Epistemic Contexts. [Preprint]

Preview

Text ProbUncert_20180524.pdf - Draft Version Download (545kB) | Preview

Abstract

This work is part of the larger project INTEGRITY. Integrity develops a conceptual frame integrating beliefs
with individual (and consensual group) decision making and action based on belief awareness. Comments and
criticisms are most welcome via email.

Starting with a thorough discussion of the conceptual embedding in existing schools of thought and liter-
ature we develop a framework that aims to be empirically adequate yet scalable to epistemic states where an
agent might testify to uncertainly believe a propositional formula based on the acceptance that a propositional
formula is possible, called accepted truth. The familiarity of human agents with probability assignments make
probabilism particularly appealing as quantitative modelling framework for defeasible reasoning that aspires
empirical adequacy for gradual belief expressed as credence functions. We employ the inner measure induced
by the probability measure, going back to Halmos, interpreted as estimate for uncertainty. Doing so omits
generally requiring direct probability assignments testi�ed as strength of belief and uncertainty by a human
agent. We provide a logical setting of the two concepts uncertain belief and accepted truth, completely relying
on the the formal frameworks of 'Reasoning about Probabilities' developed by Fagin, Halpern and Megiddo and
the 'Metaepistemic logic MEL' developed by Banerjee and Dubois. The purport of Probabilistic Uncertainty is
a framework allowing with a single quantitative concept (an inner measure induced by a probability measure)
expressing two epistemological concepts: possibilities as belief simpliciter called accepted truth, and the agents'
credence called uncertain belief for a criterion of evaluation, called rationality. The propositions accepted to be
possible form the meta-epistemic context(s) in which the agent can reason and testify uncertain belief or suspend
judgement.

---

Export/Citation:

EndNote | BibTeX | Dublin Core | ASCII/Text Citation (Chicago) | HTML Citation | OpenURL

Social Networking:

Share |

---

Item Type:	Preprint
Creators:	Creators Email ORCID Culka, M. 0000-0002-8580-9452
Keywords:	belief modeling, uncertainty, probability theory
Subjects:	General Issues > Confirmation/Induction General Issues > Determinism/Indeterminism General Issues > Formal Learning Theory General Issues > Structure of Theories
Depositing User:	Dr. M. Culka
Date Deposited:	24 May 2018 15:33
Last Modified:	24 May 2018 15:33
Item ID:	14707
Subjects:	General Issues > Confirmation/Induction General Issues > Determinism/Indeterminism General Issues > Formal Learning Theory General Issues > Structure of Theories
Date:	20 May 2018
URI:	https://philsci-archive.pitt.edu/id/eprint/14707

Available Versions of this Item

• A Probabilistic Modelling Approach for Rational Belief in Meta-Epistemic Contexts. (deposited 24 May 2018 15:33) [Currently Displayed]
  • A Probabilistic Modelling Approach for Rational Belief in Meta-Epistemic Contexts. (deposited 25 May 2018 19:24)

Monthly Views for the past 3 years

Monthly Downloads for the past 3 years

Plum Analytics

Actions (login required)

View Item