Informatik
Refine
Year of publication
Document Type
- Preprint (747)
- Article (400)
- Working Paper (119)
- Doctoral Thesis (92)
- Diploma Thesis (46)
- Conference Proceeding (41)
- Book (37)
- Bachelor Thesis (35)
- diplomthesis (29)
- Report (25)
Has Fulltext
- yes (1603)
Is part of the Bibliography
- no (1603)
Keywords
Institute
- Informatik (1603)
- Frankfurt Institute for Advanced Studies (FIAS) (1001)
- Physik (982)
- Mathematik (55)
- Präsidium (41)
- Medizin (25)
- Biowissenschaften (21)
- Exzellenzcluster Makromolekulare Komplexe (8)
- Psychologie (8)
- Deutsches Institut für Internationale Pädagogische Forschung (DIPF) (5)
Automatic termination proofs of functional programming languages are an often challenged problem Most work in this area is done on strict languages Orderings for arguments of recursive calls are generated In lazily evaluated languages arguments for functions are not necessarily evaluated to a normal form It is not a trivial task to de ne orderings on expressions that are not in normal form or that do not even have a normal form We propose a method based on an abstract reduction process that reduces up to the point when su cient ordering relations can be found The proposed method is able to nd termination proofs for lazily evaluated programs that involve non terminating subexpressions Analysis is performed on a higher order polymorphic typed language and termi nation of higher order functions can be proved too The calculus can be used to derive information on a wide range on di erent notions of termination.
Korrektur zu: C.P. Schnorr: Security of 2t-Root Identification and Signatures, Proceedings CRYPTO'96, Springer LNCS 1109, (1996), pp. 143-156 page 148, section 3, line 5 of the proof of Theorem 3. Die Korrektur wurde präsentiert als: "Factoring N via proper 2 t-Roots of 1 mod N" at Eurocrypt '97 rump session.
We analyse a continued fraction algorithm (abbreviated CFA) for arbitrary dimension n showing that it produces simultaneous diophantine approximations which are up to the factor 2^((n+2)/4) best possible. Given a real vector x=(x_1,...,x_{n-1},1) in R^n this CFA generates a sequence of vectors (p_1^(k),...,p_{n-1}^(k),q^(k)) in Z^n, k=1,2,... with increasing integers |q^{(k)}| satisfying for i=1,...,n-1 | x_i - p_i^(k)/q^(k) | <= 2^((n+2)/4) sqrt(1+x_i^2) |q^(k)|^(1+1/(n-1)) By a theorem of Dirichlet this bound is best possible in that the exponent 1+1/(n-1) can in general not be increased.
Given x small epsilon, Greek Rn an integer relation for x is a non-trivial vector m small epsilon, Greek Zn with inner product <m,x> = 0. In this paper we prove the following: Unless every NP language is recognizable in deterministic quasi-polynomial time, i.e., in time O(npoly(log n)), the ℓinfinity-shortest integer relation for a given vector x small epsilon, Greek Qn cannot be approximated in polynomial time within a factor of 2log0.5 − small gamma, Greekn, where small gamma, Greek is an arbitrarily small positive constant. This result is quasi-complementary to positive results derived from lattice basis reduction. A variant of the well-known L3-algorithm approximates for a vector x small epsilon, Greek Qn the ℓ2-shortest integer relation within a factor of 2n/2 in polynomial time. Our proof relies on recent advances in the theory of probabilistically checkable proofs, in particular on a reduction from 2-prover 1-round interactive proof-systems. The same inapproximability result is valid for finding the ℓinfinity-shortest integer solution for a homogeneous linear system of equations over Q.
We study the approximability of the following NP-complete (in their feasibility recognition forms) number theoretic optimization problems: 1. Given n numbers a1 ; : : : ; an 2 Z, find a minimum gcd set for a1 ; : : : ; an , i.e., a subset S fa1 ; : : : ; ang with minimum cardinality satisfying gcd(S) = gcd(a1 ; : : : ; an ). 2. Given n numbers a1 ; : : : ; an 2 Z, find a 1-minimum gcd multiplier for a1 ; : : : ; an , i.e., a vector x 2 Z n with minimum max 1in jx i j satisfying P n...
A partial rehabilitation of side-effecting I/O : non-determinism in non-strict functional languages
(1996)
We investigate the extension of non-strict functional languages like Haskell or Clean by a non-deterministic interaction with the external world. Using call-by-need and a natural semantics which describes the reduction of graphs, this can be done such that the Church-Rosser Theorems 1 and 2 hold. Our operational semantics is a base to recognise which particular equivalencies are preserved by program transformations. The amount of sequentialisation may be smaller than that enforced by other approaches and the programming style is closer to the common one of side-effecting programming. However, not all program transformations used by an optimising compiler for Haskell remain correct in all contexts. Our result can be interpreted as a possibility to extend current I/O-mechanism by non-deterministic deterministic memoryless function calls. For example, this permits a call to a random number generator. Adding memoryless function calls to monadic I/O is possible and has a potential to extend the Haskell I/O-system.
An anaphor resolution algorithm is presented which relies on a combination of strategies for narrowing down and selecting from antecedent sets for re exive pronouns, nonre exive pronouns, and common nouns. The work focuses on syntactic restrictions which are derived from Chomsky's Binding Theory. It is discussed how these constraints can be incorporated adequately in an anaphor resolution algorithm. Moreover, by showing that pragmatic inferences may be necessary, the limits of syntactic restrictions are elucidated.
We present a framework for the self-organized formation of high level learning by a statistical preprocessing of features. The paper focuses first on the formation of the features in the context of layers of feature processing units as a kind of resource-restricted associative multiresolution learning We clame that such an architecture must reach maturity by basic statistical proportions, optimizing the information processing capabilities of each layer. The final symbolic output is learned by pure association of features of different levels and kind of sensorial input. Finally, we also show that common error-correction learning for motor skills can be accomplished also by non-specific associative learning. Keywords: feedforward network layers, maximal information gain, restricted Hebbian learning, cellular neural nets, evolutionary associative learning
After a short introduction into traditional image transform coding, multirate systems and multiscale signal coding the paper focuses on the subject of image encoding by a neural network. Taking also noise into account a network model is proposed which not only learns the optimal localized basis functions for the transform but also learns to implement a whitening filter by multi-resolution encoding. A simulation showing the multi-resolution capabilitys concludes the contribution.
We call a vector x/spl isin/R/sup n/ highly regular if it satisfies =0 for some short, non-zero integer vector m where <...> is the inner product. We present an algorithm which given x/spl isin/R/sup n/ and /spl alpha//spl isin/N finds a highly regular nearby point x' and a short integer relation m for x'. The nearby point x' is 'good' in the sense that no short relation m~ of length less than /spl alpha//2 exists for points x~ within half the x'-distance from x. The integer relation m for x' is for random x up to an average factor 2/sup /spl alpha//2/ a shortest integer relation for x'. Our algorithm uses, for arbitrary real input x, at most O(n/sup 4/(n+log A)) many arithmetical operations on real numbers. If a is rational the algorithm operates on integers having at most O(n/sup 5/+n/sup 3/(log /spl alpha/)/sup 2/+log(/spl par/qx/spl par//sup 2/)) many bits where q is the common denominator for x.
We introduce algorithms for lattice basis reduction that are improvements of the famous L3-algorithm. If a random L3-reduced lattice basis b1,b2,...,bn is given such that the vector of reduced Gram-Schmidt coefficients ({µi,j} 1<= j< i<= n) is uniformly distributed in [0,1)n(n-1)/2, then the pruned enumeration finds with positive probability a shortest lattice vector. We demonstrate the power of these algorithms by solving random subset sum problems of arbitrary density with 74 and 82 many weights, by breaking the Chor-Rivest cryptoscheme in dimensions 103 and 151 and by breaking Damgard's hash function.
Die Anfänge der Gittertheorie reichen in das letzte Jahrhundert, wobei die wohl bekanntesten Ergebnisse auf Gauß, Hermite und Minkowski zurückgehen. Die Arbeiten sind jedoch zumeist in der Schreibweise der quadratischen Formen verfaßt, erst in den letzten Jahrzehnten hat sich die von uns verwendete Gitterschreibweise durchgesetzt. Diese ist zum einen geometrisch anschaulicher, zum anderen wurden in den letzten Jahren für diese Schreibweise effiziente Algorithmen entwickelt, so daß Probleme der Gittertheorie mittels Computer gelöst werden können. Ein wichtiges Problem ist, in einem Gitter einen kürzesten nicht verschwindenden Vektor zu bestimmen. Den Grundstein für diese algorithmische Entwicklung legten A.K. Lenstra, H.W. Lenstra Jr. und L. Lovasz mit ihrer Arbeit. In dieser führten sie einen Reduktionsbegriff ein, der durch einen Polynomialzeitalgorithmus erreicht werden kann. Ein weiterer Reduktionsbegriff, die Blockreduktion, geht auf Schnorr zurück. Euchner hat im Rahmen seiner Diplomarbeit effiziente Algorithmen für diese beiden Reduktionsbegriffe auf Workstations implementiert und auch in Dimensionen > 100 erfolgreich getestet. Die Verbesserungen von Schnittechniken des in der Blockreduktion verwendeten Aufzählungsverfahrens und die Einführung einer geschnittenen Aufzählung über die gesamte Gitterbasis hat Hörner in seiner Diplomarbeit beschrieben. Ziel der folgenden Arbeit war es nun, diese bereits auf sequentiellen Computern implementierten Algorithmen zu modifizieren, um auf parallelen Rechnern, speziell Vektorrechnern, einen möglichst hohen Geschwindigkeitsgewinn zu erzielen. Wie in den seriellen Algorithmen werden die Basisvektoren stets in exakter Darstellung mitgeführt, so daß das Endergebnis einer Berechnung nicht durch Rundungsfehler verfälscht wird.
Im Jahr 1993 schlug A. Shamir Protokolle zur Erstellung digitaler Unterschriften vor, die auf rationalen Funktionen kleinen Grades beruhen. D. Coppersmith, J. Stern und S. Vaudenay präsentierten die ersten Angriffe auf die Verfahren. Diese Angriffe können den geheimen Schlüssel nicht ermitteln. Für eine der von Shamir vorgeschlagenen Varianten zeigen wir, wie der geheime Schlüssel ermittelt werden kann. Das zweite Signaturschema von Shamir hängt von der Wahl einer algebraischen Basis ab. Eine besondere Bedeutung haben Basen, deren Elemente polynomiale Terme vom Grad 2 sind. Wir analysieren die Struktur der algebraischen Basen. Für den hervorgehobenen Spezialfall kann eine vollständige Klassifikation durchgeführt werden.
One of the most interesting domains of feedforward networks is the processing of sensor signals. There do exist some networks which extract most of the information by implementing the maximum entropy principle for Gaussian sources. This is done by transforming input patterns to the base of eigenvectors of the input autocorrelation matrix with the biggest eigenvalues. The basic building block of these networks is the linear neuron, learning with the Oja learning rule. Nevertheless, some researchers in pattern recognition theory claim that for pattern recognition and classification clustering transformations are needed which reduce the intra-class entropy. This leads to stable, reliable features and is implemented for Gaussian sources by a linear transformation using the eigenvectors with the smallest eigenvalues. In another paper (Brause 1992) it is shown that the basic building block for such a transformation can be implemented by a linear neuron using an Anti-Hebb rule and restricted weights. This paper shows the analog VLSI design for such a building block, using standard modules of multiplication and addition. The most tedious problem in this VLSI-application is the design of an analog vector normalization circuitry. It can be shown that the standard approaches of weight summation will not give the convergence to the eigenvectors for a proper feature transformation. To avoid this problem, our design differs significantly from the standard approaches by computing the real Euclidean norm. Keywords: minimum entropy, principal component analysis, VLSI, neural networks, surface approximation, cluster transformation, weight normalization circuit.
Let b1, . . . , bm 2 IRn be an arbitrary basis of lattice L that is a block Korkin Zolotarev basis with block size ¯ and let ¸i(L) denote the successive minima of lattice L. We prove that for i = 1, . . . ,m 4 i + 3 ° 2 i 1 ¯ 1 ¯ · kbik2/¸i(L)2 · ° 2m i ¯ 1 ¯ i + 3 4 where °¯ is the Hermite constant. For ¯ = 3 we establish the optimal upper bound kb1k2/¸1(L)2 · µ3 2¶m 1 2 1 and we present block Korkin Zolotarev lattice bases for which this bound is tight. We improve the Nearest Plane Algorithm of Babai (1986) using block Korkin Zolotarev bases. Given a block Korkin Zolotarev basis b1, . . . , bm with block size ¯ and x 2 L(b1, . . . , bm) a lattice point v can be found in time ¯O(¯) satisfying kx vk2 · m° 2m ¯ 1 ¯ minu2L kx uk2.