### Refine

#### Year of publication

- 1994 (40) (remove)

#### Document Type

- Preprint (40) (remove)

#### Keywords

- Boltzmann Equation (2)
- Numerical Simulation (2)
- CPLD (1)
- Case-Based Classification Algorithms (1)
- Case-Based Planning (1)
- Case-Based Representability (1)
- Domain Decomposition (1)
- Fallbasierte Planung (1)
- PABS-Methode (1)
- Particle Methods (1)

#### Faculty / Organisational entity

We present a convenient notation for positive/negativeADconditional equations. Theidea is to merge rules specifying the same function by using caseAD, ifAD, matchAD, and letADexpressions.Based on the presented macroADruleADconstruct, positive/negativeADconditional equational specifiADcations can be written on a higher level. A rewrite system translates the macroADruleADconstructsinto positive/negativeADconditional equations.

This paper presents the systematic synthesis of a fairly complex digitalcircuit and its CPLD implementation as an assemblage of communicatingasynchronous sequential circuits. The example, a VMEbus controller, waschosen because it has to control concurrent processes and to arbitrateconflicting requests.

The distribution of quasiprimary fields of fixed classes characterized by their O(N) representations Y and the number p of vector fields from which they are composed at N=infty in dependence on their normal dimension delta is shown to obey a Hardy-Ramanujan law at leading order in a 1/N-expansion. We develop a method of collective fusion of the fundamental fields which yields arbitrary qps and resolves any degeneracy.

In spite of its lack of theoretical justification, nonlinear diffusion filtering has become a powerful image enhancement tool in the recent years. The goal of the present paper is to provide a mathematical foundation for nonlinear diffusion filtering as a scale-space transformation which is flexible enough to simplify images without loosing the capability of enhancing edges. By stuying the Lyapunow functional, it is shown that nonlinear diffusion reduces Lp norms and central moments and increases the entropy of images. The proposed anisotropic class utilizes a diffusion tensor which may be adapted to the image structure. It permits existence, uniqueness and regularity results, the solution depends continuously on the initial image, and it fulfills an extremum principle. All considerations include linear and certain nonlinear isotropic models and apply to m-dimensional vector-valued images. The results are juxtaposed to linear and morphological scale-spaces.

The problem to be discussed here, is the usage of neural network clustering techniques on a mobile robot, in order to build qualitative topologic environment maps. This has to be done in realtime, i.e. the internal world model has to be adapted by the flow of sensor- samples without the possibility to stop this data-flow.Our experiments are done in a simulation environment as well as on a robot, called ALICE.

Particle Methods
(1994)

In this paper we deal with the problem of computing the stresses in stationary loaded bearings. A method to obtain the pressure in the lubrication fluid, which is given as a solution of Reynolds" differential equation, is presented. Furthermore, using the theory of plain stress, the stresses in the bearing shell are described by derivatives of biharmonic functions. A spline interpolation method for computing these functions is developed and an estimate for the error on the boundaries is presented. Finally the described methods are tested theoretically as well as with real examples.

Within the present paper we investigate case-based representability as well as case-based learnability of indexed families of uniformly recursive languages. Since we are mainly interested in case-based learning with respect to an arbitrary fixed similarity measure, case-based learnability of an indexed family requires its representability, first. We show that every indexed family is case- based representable by positive and negative cases. If only positive cases are allowed the class of representable families is comparatively small. Furthermore, we present results that provide some bounds concerning the necessary size of case bases. We study, in detail, how the choice of a case selection strategy influences the learning capabilities of a case-based learner. We define different case selection strategies and compare their learning power to one another. Furthermore, we elaborate the relations to Gold-style language learning from positive and both positive and negative examples.

Based on normalized coprime factorizations with respect to indefinite metrics and the construction of suitable characteristic functions, the Ober balanced canonical forms for the classes of bounded real and positive real are derived. This uses a matrix representation of the shift realization with respect to a basis related to sets of orthogonal polynomials.

The Boltzmann equation solutions are considered for the small Knudsen number. The main attention is devoted to certain deviations from the classical Navier-Stokes description. The equations for the quasistationary slow flows are derived. These equations do not contain the Knudsen number and provide in this sense a limiting description of hydrodynamical variables. Two well-known special cases are also indicated. In the isothermal case the equations are equivalent to the incompressible Navier-Stokes equations, in stationary case they coincide with the equations of slow non-isothermal flows. It is shown that the derived equations possess all principal properties of the Boltzmann equation on contrast to the Burnett equations. In one dimension the equations reduce to the nonlinear diffusion equations, being exactly solvable for Maxwell molecules. Multidimensional stationary heat-transfer problems are also discussed. It is shown that one can expect an essential difference between the Boltzmann equaiton solution in the limit of the continuous media and the corresponding solution of the Navier-Stokes equations.

Based on the idea of using topologic feature-mapsinstead of geometric environment maps in practical mobile robot tasks, we show an applicable way tonavigate on such topologic maps. The main features regarding this kind of navigation are: handling of very inaccurate position (and orientation) information as well as implicit modelling of complex kinematics during an adaptation phase. Due to the lack of proper a-priori knowledge, a re-inforcement based model is used for the translation of navigator commands to motor actions. Instead of employing a backpropagation network for the cen-tral associative memory module (attaching actionprobabilities to sensor situations resp. navigatorcommands) a much faster dynamic cell structure system based on dynamic feature maps is shown. Standard graph-search heuristics like A* are applied in the planning phase.

Whenever new parts of a car have been developed, the manufacturer needs an estimation of the lifetime of this new part. On one hand the construction must not be too weak, so that the part holds long enough to satisfy the customer, but on the other hand, if the construction is too excessive, the part gets too heavy.; One is interested in methods that only need few measured data from the specimen itself, but use data about the material, because constructing and testing of specimen is expensive.

Lernen von Abstraktionshierarchien zur Optimierung der Auswahl von maschinell abstrahierten Plänen
(1994)

Mit Hilfe von "Multistrategy" Ansätzen, die erklärungsbasiertes und induktives Lernen integrieren, ist es möglich, die Performanz von Planungssystemen signifikant zu verbessern. Dabei können gelöste Planungsprobleme zunächst mit einem wissensintensiven Verfahren abstrahiert und generalisiert werden. Durch den in diesem Beitrag im Vordergrund stehenden induktiven inkrementellen Lernalgorithmus ist es dann weiterhin möglich, die Gesamtheit des deduktiv generierten Wissens in einer Abstraktionshierarchie anzuordnen. Dabei wird die, im allgemeinen unentscheidbare, "spezieller-als-Relation" zwischen generalisierten Plänen, induktiv aus den gegebenen Planungsfällen gelernt. Diese Abstraktionshierarchie dient dann zur Klassifikation neuer Problemstellungen und damit zur Bestimmung einer speziellsten anwendbaren abstrakten Problemlösung.

While symbolic learning approaches encode the knowledge provided by the presentation of the cases explicitly into a symbolic representation of the concept, e.g. formulas, rules, or decision trees, case-based approaches describe learned concepts implicitly by a pair (CB; d), i.e. by a set CB of cases and a distance measure d. Given the same information, symbolic as well as the case-based approach compute a classification when a new case is presented. This poses the question if there are any differences concerning the learning power of the two approaches. In this work we will study the relationship between the case base, the measure of distance, and the target concept of the learning process. To do so, we transform a simple symbolic learning algorithm (the version space algorithm) into an equivalent case-based variant. The achieved results strengthen the conjecture of the equivalence of the learning power of symbolic and casebased methods and show the interdependency between the measure used by a case-based algorithm and the target concept.

Abstract: We calculate exact analytical expressions for O(alpha s) 3-jet and O (alpha^2 s ) 4-jet cross sections in polarized deep inelastic lepton nucleon scattering. Introducing an invariant jet definition scheme, we present differential distributions of 3- and 4-jet cross sections in the basic kinematical variables x and W^2 as well as total jet cross sections and show their dependence on the chosen spin-dependent (polarized) parton distributions. Noticebly differences in the predictions are found for the two extreme choices, i.e. a large negative sea-quark density or a large positive gluon density. Therefore, it may be possible to discriminate between different parametrizations of polarized parton densities, and hence between the different physical pictures of the proton spin underlying these parametrizations.