摘要:

In the experiments of pill bugs (Armadillidium vulgare), some of them performed as if they had contained the models of themselves and the environment in view of computing their present state as a function of the prediction of the models. In a specific situation, they escaped from the experimental apparatus as if they had constituted spatial knowledge of it (open field surrounded by walls) in the process of exploratory behavior and used the knowledge. This species gets environmental information by tactile ability of antennae, not by visual one, and do not climb perpendicular walls in general condition. If they had not escaped in the experiment, they would have died of hunger or water deficit. In this paper I will present the result of this anticipatory behavior. I also discuss that the notion of anticipation, which is another name of autonomy, is inevitably introduced when one considers the process of understanding of animal behavior progressing without any common basis between animals and experimenters. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58244

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

The representation problem, which plays a crucial role in the study of cognition, is discussed from a holistic perspective. Based on some distinctions a concept of duality is introduced, which can be applied to two aspects involved in the representation problem, namely meaning and computation. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58261

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

This paper aims at laying down the foundations of a new approach to learning in autonomous mobile robots. It is based on the assumption that robots can be provided with built-in action plans and with mechanisms to modify and improve such plans. This requires that robots are equipped with some form of high-level reasoning capabilities. Therefore, the proposed learning technique is embedded in a novel distributed control architecture featuring an explicit model of robot’s cognitive activity. In particular, cognitive activity is obtained by the interaction ofactive mental entities, such as intentions, persuasions and expectations. Learning capabilities are implemented starting from the interaction of such mental entities. The proposal is illustrated through an example concerning a robot in charge of reaching a target in an unknown environment cluttered with obstacles. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58262

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

We study changes of coordinates that allow the representation of the ordinary differential equations describing continuous-time recurrent neural networks into differential equations describing predator-prey models—also called Lotka-Volterra systems. We transform the equations for the neural network first into quasi-monomial form, where we express the vector field of the dynamical system as a linear combination of products of powers of the variables. In practice, this transformation is possible only if the activation function is the hyperbolic tangent or the logistic sigmoı¨d. From this quasi-monomial form, we can directly transform the system further into Lotka-Volterra equations. The resulting Lotka-Volterra system is of higher dimension than the original system, but the behavior of its first variables is equivalent to the behavior of the original neural network. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58279

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

We discuss a method to implement in a Boolean neural network a Hebbian rule so to obtain an adaptable universal control system. We start by presenting both the Boolean neural net and the Hebbian rule we have considered. Then we discuss, first, the problems arising when the latter is naively implemented in a Boolean neural net, second, the method consenting us to overcome them and the ensuing adaptable Boolean neural net paradigm. Next, we present the adaptable Boolean neural net as an intelligent control system, actually controlling a writing robot, and discuss how to train it in the execution of the elementary arithmetic operations on operands represented by numerals with an arbitrary number of digits. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58263

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

A categorical semantic domain is constructed for Petri nets which satisfies the diagonal compositionality requirement with respect to anticipations, i.e., Petri nets are equipped with a compositional anticipation mechanism (vertical compositionality) that distributes through net combinators (horizontal compositionality). The anticipation mechanism is based on graph transformations (single pushout approach). A finitely bicomplete category of partial Petri nets and partial morphisms is introduced. Classes of transformations stand for anticipations. The composition of anticipations (i.e., composition of pushouts) is defined, leading to a category of nets and anticipations which is also complete and cocomplete. Since the anticipation operation composes, the vertical compositionality requirement of Petri nets is achieved. Then, it is proven that the anticipation also satisfies the horizontal compositionality requirement. A specification grammar stands for a system specification and the corresponding induced subcategory of nets and anticipation’s stands for all possible dynamic anticipation’s of the system (objects) and their relationship (morphims). ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58245

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

Basic biological information is stored in strings of nucleic acids (DNA, RNA) or amino acids (proteins). Teasing out the meaning of these strings is a central problem of modern biology. Matching and aligning strings brings out their shared characteristics. Although string matching is well-understood in the edit-distance model, biological strings with transpositions and inversions violate this model’s assumptions. We propose a family of heuristics calledwalking treesto align biologically reasonable strings. Both edit-distance and walking tree methods can locate specific genes within a large string when the genes’ sequences are given. When we attempt to match whole strings, the walking tree matches most genes, while the edit-distance method fails. We also give examples in which the walking tree matches substrings even if they have been moved or inverted. The edit-distance method was not designed to handle these problems. We include an example in which the walking tree “discovered” a gene. Calculating scores for whole genome matches gives a method for approximating evolutionary distance. We show two evolutionary trees for the picornaviruses which were computed by the walking tree heuristic. Both of these trees show great similarity to previously constructed trees. The point of this demonstration is that WHOLE genomes can be matched and distances calculated. The first tree was created on a Sequent parallel computer and demonstrates that the walking tree heuristic can be efficiently parallelized. The second tree was created using a network of work stations and demonstrates that there is suffient parallelism in the phylogenetic tree calculation that the sequential walking tree can be used effectively on a network. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58246

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

Genetic Algorithms (GAs) are adaptive search methods, which try to incorporate the principle of surviving known from nature. They proved to be an efficient instrument for solving many hard problems in different areas, in which the majority of other techniques failed as being weak or not applicable. On the other hand, GAs fight with a number of problems, as well. To the crucial issues belong the representation of potential solutions in the search space, design of the proper operators that drive the search and the configuration of the routines and strategies used in the GAs. This paper presents some results of our research on GAs. The interesting observations encountered by experiments concerning the initialisation of GAs’ runs and by the enhanced crossover operator for binary chromosomes are presented. We have used a shell calledGAToolfor automatic experimenting with GAs, which was developed for GAs’ performance evaluation. The representatives of Travelling Salesman Problem (TSP), Continuous functions and Deceptive functions, which are the usual benchmarks mentioned in many studies, were used in our experiments. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58247

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

We consider the problem of discovering Datalog rules with negation that aresignificant(i.e. interesting andsufficiently valid) with respect to a given set of positive and negative facts. We propose a formal framework, as well as algorithms, for the discovery of such rules from a set of facts that are stored in the form of a set of tables in a database. Additionally, we propose a way of updating the discovered rules when the set of facts is updated. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58280

出版商:AIP    

年代:1999

数据来源: AIP

摘要:

With the growing amount of distributed and heterogeneous information and services, conventional information systems have come to their limits. This gave rise to the development of a Multi-Agent System (the “Logical Client”) which can be used in complex information systems as well as in other advanced software systems. Computer agents are proactive, reactive and social. They form a community of independent software components that can communicate and co-operate in order to accomplish complex tasks. Thus the agent-oriented paradigm provides a new and powerful approach to programming distributed systems. The communication framework developed is based on standards like CORBA, KQML and KIF. It provides an embedded rule based system to find adequate reactions to incoming messages. The macro-architecture of the Logical Client consists of independent agents and uses artificial intelligence to cope with complex patterns of communication and actions. A set of system agents is also provided, including the Strategy Service as a core component for modelling processes at runtime, the Computer Supported Cooperative Work (CSCW) Component for supporting remote co-operation between human users and the Repository for managing and hiding the file based data flow in heterogeneous networks. This architecture seems to be capable of managing complexity in information systems. It is also being implemented in a complex simulation system that monitors and simulates the environmental radioactivity in the country Baden-Wu¨rttemberg. ©1999 American Institute of Physics.

ISSN:0094-243X    

DOI:10.1063/1.58248

出版商:AIP    

年代:1999

数据来源: AIP