Computational Intelligence - August 2012 - 12

function evaluations. Particle Swarm
Optimization combined with Gradient
Descent starting from a small number of
a-cuts leads to the most accurate fuzzy
intervals at the cost of a relatively large
number of function evaluations. The
authors provide that the best approach
to construct the membership function is
to use the Fuzzy Calculator with an
increasing number of a-cuts, with PSO
GD as the optimization algorithm, using
a population size of 20 particles and
communication at every ﬁve iterations
and by recalculating the optima if
inconsistencies between subsequent
a-cuts occur."
A Linguistic Approach to Influencing
Decision Behavior, by F.E. Petry and
R.R. Yager, IEEE Transactions on
Fuzzy Systems, Vol. 20, No. 2, Apr.
2012, pp. 248-261.
Digital Object Identifier: 10.1109/
TFUZZ.2011.2172795
"In this paper, the authors provide a
set of approach to influence human's
decision-making behavior-to draw a
conclusion "V is P" given "V is F,"
where P is a fuzzy subset of F representing some linguistic value for V that corresponds to a perception of the world V
is P, which they want a person to accept.
The authors examine several methods to
represent the process to engender such
influences. These include using a person's predispositions, framing the context of a discussion and generalization
techniques that allow issues to be
viewed in a more favorable light. These
extensions can provide closer approximations to the uncertainty found in
human persuasion behavior."
IEEE Transactions on
Evolutionary Computation

Adaptation in Dynamic Environments:
A Case Study in Mission Planning, by
L. Bui, Z. Michalewicz, E. Parkinson,
and M. Abello, IEEE Transactions on

Evolutionary Computation, Vol. 16,
No. 2, Apr. 2012, pp. 190-209.
Digital Object Identifier: 10.1109/
TEVC.2010.2104156
"Random events are usually associated with execution of operational
plans in companies and organizations.
This situation poses problems for planning staff who must generate plans
that can adapt as needed. This paper
deals with adaption in dynamic environments. The dynamic planning
problem is cast as a multiobjective
optimization problem and the proposed method is used to produce a
non-dominated solution set after each
planning cycle. A mission planning
case study is used to show the benefit
of the authors' approach."
Evolving Distributed Algorithms with
Genetic Programming, by T. Weise and
K. Tang, IEEE Transactions on Evolutionary Computation, Vol. 16, No. 2,
Apr. 2012, pp. 242-265.
Digital Object Identifier: 10.1109/
TEVC.2011.2112666
"This paper evaluates the applicability of genetic programming for the evolution of distributed algorithms. A
simulation environment is constructed
to model various asynchronous computation phenomena such as out-of-order
message delivery. Extensions to standard
genetic programming approaches are
then described to make it suitable for
investigating distributed algorithms.
Experimental results led the authors to
conclude genetic programming is a viable method for evolving non-trivial,
deterministic, non-approximative distributed algorithms."
IEEE Transactions on
Computational Intelligence
and AI in Games

A Robust Learning Approach to Repeated Auctions with Monitoring and Entry

IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE | AUGUST 2012

Fees, by A. Danak and S. Mannor,
IEEE Transactions on Computational
Intelligence and AI in Games, Vol. 3,
No. 4, Dec. 2011, pp. 302-315.
Digital Object Identifier: 10.1109/
TCIAIG.2011.2160994
"The paper tackles a realistic auction model and shows how a simple
learning strategy can make an efficient
estimation of market conditions and
use such estimates to make informed
bids. The strategies are proven to be
stable and robust, and shown to perform well in a search engine marketing experiment."
IEEE Transactions on Autonomous
Mental Development

Autonomous Learning of High-Level
States and Actions in Continuous Environments, by J. Mugan and B. Kuipers,
IEEE Transactions on Autonomous Mental Development, Vol. 4, No. 1, Mar.
2012, pp. 70-86.
Digital Object Identifier: 10.1109/
TAMD.2011.2160943
"This paper attacks the problem of
learning high-level states and actions in
continuous environments by using a
qualitative representation to bridge the
gap between continuous and discrete
variable representations. In this approach,
the agent begins with a broad discretization and initially can only tell if the
value of each variable is increasing,
decreasing, or remaining steady. The
agent then simultaneously learns a qualitative representation (discretization) and
a set of predictive models of the environment. These models are converted
into plans to perform actions. The agent
then uses those learned actions to
explore the environment. The method is
evaluated using a simulated robot with
realistic physics."

Table of Contents for the Digital Edition of Computational Intelligence - August 2012