Checking Trustworthiness of Probabilistic Computations in a Typed Natural Deduction System

06/26/2022
by   Fabio Aurelio D'Asaro, et al.
0

In this paper we present the probabilistic typed natural deduction calculus TPTND, designed to reason about and derive trustworthiness properties of probabilistic computational processes, like those underlying current AI applications. Derivability in TPTND is interpreted as the process of extracting n samples of outputs with a certain frequency from a given categorical distribution. We formalize trust within our framework as a form of hypothesis testing on the distance between such frequency and the intended probability. The main advantage of the calculus is to render such notion of trustworthiness checkable. We present the proof-theoretic semantics of TPTND and illustrate structural and metatheoretical properties, with particular focus on safety. We motivate its use in the verification of algorithms for automatic classification.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
02/02/2023

A Typed Lambda-Calculus for Establishing Trust in Probabilistic Programs

The extensive deployment of probabilistic algorithms has radically chang...
research
03/02/2021

Natural deduction for intuitionistic belief: proof theory and proof-theoretic semantics

Intuitionistic belief has been axiomatized by Artemov and Protopopescu a...
research
06/20/2016

Introducing a Calculus of Effects and Handlers for Natural Language Semantics

In compositional model-theoretic semantics, researchers assemble truth-c...
research
04/18/2019

Realizability in the Unitary Sphere

In this paper we present a semantics for a linear algebraic lambda-calcu...
research
01/19/2019

Kantorovich Continuity of Probabilistic Programs

The Kantorovich metric is a canonical lifting of a distance from sets to...
research
10/29/2019

A Graph-Based Tool to Embed the π-Calculus into a Computational DPO Framework

Graph transformation approaches have been successfully used to analyse a...

Please sign up or login with your details

Forgot password? Click here to reset