Turku Centre for Computer Science

About: Turku Centre for Computer Science is a facility organization based out in Turku, Finland. It is known for research contribution in the topics: Decidability & Word (group theory). The organization has 382 authors who have published 1027 publications receiving 19560 citations.

The non-parametrizability of the word equation xyz = zvx : a short proof

Abstract: Although Makanin proved the problem of satisfiability of word equations to be decidable, the general structure of solutions is difficult to describe. In particular, Hmelevskii proved that the set of solutions of xyz = zvx cannot be described using only finitely many parameters, contrary to the case of equations in three unknowns. In this paper we give a short, elementary proof of Hmelevskii's result.

Fast retraining of artificial neural networks

Abstract: In this paper we propose a practical mechanism for extracting information directly from the weights of a reference artificial neural network (ANN). We use this information to train a structurally identical ANN that has some variations of the global transformation input-output function. To be able to fulfill our goal, we reduce the reference network weights by a scaling factor. The evaluation of the computing effort involved in the retraining of some ANNs shows us that a good choice for the scaling factor can substantially reduce the number of training cycles independent of the learning methods. The retraining mechanism is analyzed for the feedforward ANNs with two inputs and one output.

Electrophoretic signal comparison applied to mRNA differential display analysis.

Abstract: Gene expression analysis by electrophoretic methods is currently limited by the labor-intensive visual evaluation of the electrophoretic signal profiles. For this purpose, we present a flexible approach to computer-assisted comparison of quantitative electrophoretic patterns between multiple expression signals. Gaussian curves are first fitted to the complex peak mixtures, and the resulting approximate signals are then aligned and compared on a peak-by-peak basis with respect to specific patterns defined by the investigator. The rationale of the method is to produce a compressed list of exceptional expression patterns quantified by a set of associated numeric features. A score value is attached to each pattern in such a way that large values identify the most potential findings to be focused on in visual analysis instead of the vast amount of original electrophoretic results. The validity of the method is demonstrated by analyzing a large set of electrophoretic data from mRNA differential display experiments monitoring changes in gene expression patterns in human colonic carcinoma. The automated identification of variously defined gene expression patterns agrees well with the visual evaluation of the same electropherograms. The general comparison approach may also be found useful with other gene expression profiling instruments.

On algebraic decoding of the Z/sub 4/-linear Goethals-like codes

Abstract: The Z/sub 4/-linear Goethals-like code of length 2/sup m/ has 2/sup 2m+1-3m-2/ codewords and minimum Lee distance 8 for any odd integer m/spl ges/3. We present an algebraic decoding algorithm for all Z/sub 4/-linear Goethals-like codes C/sub k/ introduced by Helleseth et al.(1995, 1996). We use Dickson polynomials and their properties to solve the syndrome equations.