Logic meets Probability: Towards Explainable AI Systems for Uncertain Worlds
نویسنده
چکیده
Logical AI is concerned with formal languages to represent and reason with qualitative specifications; statistical AI is concerned with learning quantitative specifications from data. To combine the strengths of these two camps, there has been exciting recent progress on unifying logic and probability. We review the many guises for this union, while emphasizing the need for a formal language to represent a system’s knowledge. Formal languages allow their internal properties to be robustly scrutinized, can be augmented by adding new knowledge, and are amenable to abstractions, all of which are vital to the design of intelligent systems that are explainable and interpretable.
منابع مشابه
Towards a Universal Theory of Artiicial Intelligence Based on Algorithmic Probability and Sequential Decision Theory
Decision theory formally solves the problem of rational agents in uncertain worlds if the true environmental probability distribution is known. Solomonoo's theory of universal induction formally solves the problem of sequence prediction for unknown distribution. We unify both theories and give strong arguments that the resulting universal AI model behaves optimal in any computable environment. ...
متن کاملProbabilistic Belief Logics
Modal logics based on Kripke style semantics are the prominent formalism in AI for modeling beliefs Kripke semantics involve a collection of possible worlds and a relation among the worlds called an accessibility relation Dependent on the properties of the accessibility relation di erent modal operators can be captured Belief operators have been modeled by an accessibility relation which produc...
متن کاملMorph Gentzen Plan Computation
New planning techniques with model diagrams and applications to computable models with morph Gentzen computing on relevant worlds are presented. Generic diagrams are applied to model computing with localized minimal efficient computable KR on AI worlds. Diagrammatic reasoning is defined in terms of inferences directed by the G-diagrams for models. G-diagrams are applied towards KR from planning...
متن کاملWhat Does Explainable AI Really Mean? A New Conceptualization of Perspectives
We characterize three notions of explainable AI that cut across research fields: opaque systems that offer no insight into its algorithmic mechanisms; interpretable systems where users can mathematically analyze its algorithmic mechanisms; and comprehensible systems that emit symbols enabling user-driven explanations of how a conclusion is reached. The paper is motivated by a corpus analysis of...
متن کاملThe New Empiricism and the Semantic Web: Threat or Opportunity?
The early AI success stories of the 1970s were based in small ’worlds’ with carefully bounded semantic domains: Winograd’s SHRDLU (WIN72) is perhaps the canonical example. The rapid growth of efforts to found the next generation of systems on general-purpose knowledge representation languages (I’m thinking of several varieties of semantic nets, from plain to partitioned, as well as KRL, KL-ONE ...
متن کامل