Skip to main content

2004 | Buch

Intelligent Agents

Theory and Applications

verfasst von: Professor Germano Resconi, Professor Lakhmi C. Jain

Verlag: Springer Berlin Heidelberg

Buchreihe : Studies in Fuzziness and Soft Computing

insite
SUCHEN

Über dieses Buch

This research book presents the agent theory and adaptation of agents in different contexts. Agents of different orders of complexity must be autonomous in the rules used. The agent must have a brain by which it can discover rules contained within the data. Because rules are the instruments by which agents change the environment, any adaptation of the rules can be considered as an evolution of the agents. Because uncertainty is present in every context, we shall describe in this book how it is possible to introduce global uncertainty from the local world into the description of the rules. This book contains ten chapters. Chapter 1 gives a general dscription of the evolutionary adaptation agent. Chapter 2 describes the actions and meta­ actions of the agent at different orders. Chapter 3 presents in an abstract and formal way the actions at different orders. Chapter 4 connects systems and meta systems with the adaptive agent. Chapter 5 describes the brain of the agent by the morphogenetic neuron theory. Chapter 6 introduces the logic structure of the adaptive agent. Chapter 7 describes the feedback and hyper-feedback in the adaptive agent. Chapter 8 introduces the adaptation field into the modal logic space as logic instrument in the adaptive agent. Chapter 9 describes the action of the agent in the physical domain. Chapter 10 presents the practical application of agents in robots and evolutionary computing.

Inhaltsverzeichnis

Frontmatter
Chapter One. Evolutionary Adaptive Agents
Abstract
The subject of this book is the agent theory or Adaptive Agents in different contexts. Agents of different orders of complexity must be autonomous in the rules used. The agent must have a brain by which it can discover the rules contained within the data. Because rules are the instruments by which agents change the environment, any adaptation of the rules can be considered as an evolution of the agents. As uncertainty is present in every context, we shall describe in this book how to introduce global uncertainty from the local world into the description of the rules. In conclusion adaptation of different orders, uncertainty and the rules obtained by the data, form a new entity that can simulate different natural phenomena.
Germano Resconi, Lakhmi C. Jain
Chapter Two. Adaptive Agents and Their Actions
Abstract
In this chapter we present the fundamental mathematical and cognitive scheme in which we locate the Adaptive Agent. We introduce the context definition at different orders, the transformations from one context to another and the adaptation processes by which we can establish a symmetric relation among the rules inside the different contexts. This chapter is not completely abstract but has the aim to introduce the concept of the Adaptive Agent. When we have only one context, adaptation is impossible. The definition of every context is made by variables and rules and the rules are fixed inside the context. The adaptation principle has its source within the change of the rules. Consequently because we have only one context and every change of the context is forbidden, it is not possible to have any adaptation process.
Germano Resconi, Lakhmi C. Jain
Chapter Three. Abstract Theory of Adaptive Agents
Abstract
The syntactic and semantic aspects of the formal language of General System Logical Theory (GSLT), also known as Abstract Theory of Adaptive Agents is presented in order to improve the formal description of the action of the adaptive agents. Following an introduction to the language of GSLT, we show the possibility of the action of the Adaptive Agents in a number of areas. The category theory will be the abstract structure to model the adaptive agent at different orders. Every semantic unity which describes an action of the adaptive agent is expressed by an elementary unity denoted Elementary Logical Systems (ELS) which are the basic components of GSLT. GSLT uses the input-output paradigm to represent any action. The static classical form of categorical structures is transformed as a dynamic transformation of objects. The input-output paradigm of system theory is well known and is extensively developed in the literature.
Germano Resconi, Lakhmi C. Jain
Chapter Four. Adaptive Agents and Complex Systems
Abstract
In the definition given by G. Klir, a system S is defined as S = [M, R] where M is a set of objects and R is a set of relations between the objects. We can use this system definition to describe the Logic where the objects are propositions and the relations are the inferential rules. We can also describe the semantics of a linguistic text (semantic web) and many other disciplines. In spite of the success of the system model, we argue that a higher order of systems exists. The systems of order one are ordinary systems with objects and relations. The systems of order two are meta-systems where the relations are transformed into one another. Every transformation changes the relations which are in one context into the image of the same relations in another context. Transformations of relations or functions can be homomorphisms.
Germano Resconi, Lakhmi C. Jain
Chapter Five. Adaptive Agents and Models of the Brain Functions
Abstract
In the previous chapters we have presented the action of the adaptive agent at different orders with its abstract and functional image. Now we give a model of the action that is considered to be similar to the brain functions. We then argue that the brain is a model of the action of the adaptive agent. In an important paper Alexandre Pouger and Lawrence H. Snyder [14] use the basis functions as the main frame to model brain functions. In the abstract of the paper they state:
“Behaviours such as sensing an object and then moving your eyes or hand toward it require that sensory information be used to help generate a motor command, a process known as a sensorimotor transformation. Here we review models of sensorimotor transformations that use a flexible intermediate representation that relies on basis function. The use of basis function as an intermediate is borrowed from the theory of non linear function approximation. We show that this approach provides a unifying insight into the neural basis of three crucial aspects of sensorimotor transformations, namely, computation, learning and short term memory. This mathematical formalism is consistent with the responses of cortical neurons and provides a fresh perspective on the issue of frames of reference in spatial representations”.
Germano Resconi, Lakhmi C. Jain
Chapter Six. Logic Actions of Adaptive Agents
Abstract
There is a general acceptance that Kripke models (Chellas, 1980) [1] present the fundamental semantics for modal logic. There are several generalisations of Kripke models. Among these generalisations there are the so-called Scott-Montague models (Chellas, 1980). These have the most general characteristics of such models. In this section we present a brief description of Kripke and Scott-Montague models.
Germano Resconi, Lakhmi C. Jain
Chapter Seven. The Hierarchical Structure of Adaptive Agents
Abstract
An important definition found in the System Theory (A. Wayne Wymore, 1993) [6] is the System Coupling:
Germano Resconi, Lakhmi C. Jain
Chapter Eight. The Adaptive Field in Logical Conceptual Space
Abstract
With the definition of the Adaptive Field we may assign a particular degree of significance to any possible world in modal logic. In this way, our attention can be focused on more significant worlds or on the comparison between two different worlds. The information is one of the main sources able to show the logic structure of information. A guide is obtained from this on the use of information. It is also able to discover and measure the degree of uncertainty contained. The space of worlds is thus useful in its ability to divide information into the important parts. An accessibility relation exists between the most significant worlds with respect to the least significant worlds. Logical models of information are possible using the adaptive fields.
Germano Resconi, Lakhmi C. Jain
Chapter Nine. Adaptive Agents in the Physical Domain
Abstract
This chapter presents the study of Physical Phenomena in line with the adaptive agent described in the previous chapters. We have demonstrated that in the model of the physical reality we can find the same logical structure as that used for more conceptual and abstract fields of research.
Germano Resconi, Lakhmi C. Jain
Chapter Ten. Practical Applications of Agents in Robots and Evolution Population
Abstract
In the robot SCARA [John J.Craig, 1955] [1] the equation that connects the work space (x,y) with the joint space (α,β) is:
$$ \left\{ \begin{gathered} X = {\psi _1} = \cos \left( \alpha \right) + \cos \left( {\alpha + \beta } \right) \hfill \\ Y = {\psi _2} = \sin \left( \alpha \right) + \sin \left( {\alpha + \beta } \right) \hfill \\ \end{gathered} \right. $$
(10.1)
These positions are represented in Figure.10.1 where we show the mechanical system of the robot SCARA.
Germano Resconi, Lakhmi C. Jain
Backmatter
Metadaten
Titel
Intelligent Agents
verfasst von
Professor Germano Resconi
Professor Lakhmi C. Jain
Copyright-Jahr
2004
Verlag
Springer Berlin Heidelberg
Electronic ISBN
978-3-540-44401-5
Print ISBN
978-3-642-06031-1
DOI
https://doi.org/10.1007/978-3-540-44401-5