Decision Making Under Uncertainty: Theory and Application by Mykel J. Kochenderfer et al.

By Mykel J. Kochenderfer et al.

Many vital difficulties contain choice making below uncertainty -- that's, picking activities in accordance with frequently imperfect observations, with unknown results. Designers of computerized selection help platforms needs to consider a few of the resources of uncertainty whereas balancing the a number of goals of the method. This publication offers an advent to the demanding situations of choice making below uncertainty from a computational standpoint. It offers either the speculation in the back of choice making versions and algorithms and a set of instance purposes that diversity from speech popularity to plane collision avoidance.

Focusing on tools for designing selection brokers, making plans and reinforcement studying, the publication covers probabilistic versions, introducing Bayesian networks as a graphical version that captures probabilistic relationships among variables; software concept as a framework for realizing optimum selection making lower than uncertainty; Markov choice tactics as a mode for modeling sequential difficulties; version uncertainty; country uncertainty; and cooperative selection making related to a number of interacting brokers. a chain of purposes exhibits how the theoretical techniques might be utilized to structures for attribute-based individual seek, speech functions, collision avoidance, and unmanned plane chronic surveillance.

Decision Making less than Uncertainty unifies examine from varied groups utilizing constant notation, and is out there to scholars and researchers throughout engineering disciplines who've a few earlier publicity to chance conception and calculus. it may be used as a textual content for complicated undergraduate and graduate scholars in fields together with desktop technological know-how, aerospace and electric engineering, and administration technological know-how. it is going to even be a necessary specialist reference for researchers in various disciplines.

Show description

Read Online or Download Decision Making Under Uncertainty: Theory and Application PDF

Similar intelligence & semantics books

Cambrian Intelligence: The Early History of the New AI

This e-book is a suite of the "best" / such a lot stated Brooks papers. essentially it covers what's thought of the center of papers that bought behaviour established robotics rolling. just about all papers have seemed as magazine papers prior and this can be purely a handy number of those. For a person engaged on cellular robotics those papers are a needs to.

Recent Advances in AI Planning: 4th European Conference on Planning, ECP'97, Toulouse, France, September 24 - 26, 1997, Proceedings

This ebook constitutes the refereed court cases of the 4th eu convention on making plans, ECP'97, held in Toulouse, France, in September 1997. The 35 revised complete papers offered have been conscientiously reviewed and chosen from ninety submissions. the variety of subject matters lined spans all features of present synthetic intelligence making plans, from theoretical and foundational concerns to real making plans of platforms and functions in a number of components.

Artificial Intelligence: Its Scope and Limits

This sequence will contain monographs and collections of stories dedicated to the research and exploration of information, details, and knowledge­ processing platforms of all types, regardless of even if human, (other) animal, or desktop. Its scope is meant to span the whole diversity of pursuits from classical difficulties within the philosophy of brain and philosophical psycholo­ gy via matters in cognitive psychology and sociobiology (concerning the psychological features of alternative species) to rules on the topic of man made in­ telligence and to computing device technology.

Application of Evolutionary Algorithms for Multi-objective Optimization in VLSI and Embedded Systems

This publication describes how evolutionary algorithms (EA), together with genetic algorithms (GA) and particle swarm optimization (PSO) can be used for fixing multi-objective optimization difficulties within the region of embedded and VLSI method layout. Many complicated engineering optimization difficulties should be modelled as multi-objective formulations.

Extra resources for Decision Making Under Uncertainty: Theory and Application

Sample text

We want to capture the property that our radar is more likely to detect aircraft with larger cross sections. Of course, we could just set a threshold θ and say P (d 1 | c ) = 0 if c < θ and P (d 1 | c ) = 1 otherwise. However, such a model would potentially assign zero probability to detections that may actually occur. Instead of a hard threshold to define P (D | C ), we could use a soft threshold that assigns low probabilities when below a threshold and high probabilities above a threshold. One way to represent a soft threshold is to use a logit model, which produces a sigmoid curve having an “S” shape: P (d 1 | c ) = 1 1 + exp −2 c −θ1 θ2 .

The observation at time t is denoted O t . The observation nodes are shaded to indicate that the values at those nodes are known. If the states correspond to the position and velocity of an aircraft, then the observations may be noisy radar measurements of the range and azimuth. If the state variables are discrete, then the model is called a hidden Markov model (HMM). If the state variables are continuous and the conditional distributions are linear Gaussian, then the model is called a linear dynamical system.

Because military aircraft are more likely to be designed to have lower radar cross section than nonmilitary aircraft, we would want θ4 to be smaller than θ1 . Finally, we need to define the conditional distribution P (D | C ). We want to capture the property that our radar is more likely to detect aircraft with larger cross sections. Of course, we could just set a threshold θ and say P (d 1 | c ) = 0 if c < θ and P (d 1 | c ) = 1 otherwise. However, such a model would potentially assign zero probability to detections that may actually occur.

Download PDF sample

Rated 4.10 of 5 – based on 35 votes