Decision Making with Partially Consonant Belief Functions Artificial Intelligence

This paper studies decision making for Walley's partially consonant belief functions (pcb). In a pcb, the set of foci are partitioned. Within each partition, the foci are nested. The pcb class includes probability functions and possibility functions as extreme cases. Unlike earlier proposals for a decision theory with belief functions, we employ an axiomatic approach. We adopt an axiom system similar in spirit to von Neumann - Morgenstern's linear utility theory for a preference relation on pcb lotteries. We prove a representation theorem for this relation. Utility for a pcb lottery is a combination of linear utility for probabilistic lottery and binary utility for possibilistic lottery.

Hybrid Influence Diagrams Using Mixtures of Truncated Exponentials Artificial Intelligence

Mixtures of truncated exponentials (MTE) potentials are an alternative to discretization for representing continuous chance variables in influence diagrams. Also, MTE potentials can be used to approximate utility functions. This paper introduces MTE influence diagrams, which can represent decision problems without restrictions on the relationships between continuous and discrete chance variables, without limitations on the distributions of continuous chance variables, and without limitations on the nature of the utility functions. In MTE influence diagrams, all probability distributions and the joint utility function (or its multiplicative factors) are represented by MTE potentials and decision nodes are assumed to have discrete state spaces. MTE influence diagrams are solved by variable elimination using a fusion algorithm.

Experimentation Guided by A Knowledge Graph

AAAI Conferences

Department of Computer Science Wichita State University, Wichita, KS 67260 U.S.A. Introduction Discoverers always seek the unknown. They examine the world around them, and ask: what are the boundaries that separate the known from the unknown? Then they cross the boundaries to explore the world beyond. Machine discoverers can use the same strategy. We will discuss a knowledge representation mechanism that makes it easy to find the unknown.

KSU Willie in Scavenger Hunt at AAAI '06

AAAI Conferences

Aaron Chavez, Michael Marlen, Chris Meyer, Andrew King, Joseph Lutz, and Dr David Gustafson Computing and Information Sciences Kansas State University Manhattan, KS 66506 Introduction The Kansas State University entry into the AAAI 2006 robot scavenger hunt consists of a Pioneer P3AT robot (see figure 1) running Windows 2000, scalable client/server software architecture, blob-based object recognition, and a path-planning package coupled with an off-the-shelf Monte-Carlo package that we have augmented. As the state of the robot changes (either from default behavior coded into the server, or by other clients requesting a state change) the server generates a series of messages which are broadcast to all connected clients to notify them of the state change. All state change requests from clients are served in order of the time they were received. Figure 1: KSU's P3AT robot Client / Server Architecture For flexibility in development, we built a client / server model (see figure 2) to abstract the ARIA (ActivMedia) API and allow for distributed computation. The server process(es) run on the robot(s).

KSU Willie in the AAAI 2007 Semantic Vision Challenge

AAAI Conferences

Kansas State University competed in both the robot division and the software division of the AAAI 2007 Semantic Vision Challenge. The team used a Pioneer P3AT robot, scalable client/server software architecture, path-planning code, and a set of image classifiers that autonomously trained on images downloaded from the internet. The team succeeded in identifying multiple objects in the environment.