DSpace Collection: Volume 1 Number 1
http://hdl.handle.net/10525/316
The Collection's search engineSearch the Channelsearch
http://sci-gems.math.bas.bg/jspui/simple-search
Applying A Normalized Compression Metric To The Measurement Of Dialect Distance
http://hdl.handle.net/10525/334
Title: Applying A Normalized Compression Metric To The Measurement Of Dialect Distance<br/><br/>Authors: Simov, Kiril; Osenova, Petya<br/><br/>Abstract: The paper discusses the application of a similarity metric basedon compression to the measurement of the distance among Bulgarian dia-lects. The similarity metric is de ned on the basis of the notion of Kolmo-gorov complexity of a le (or binary string). The application of Kolmogorovcomplexity in practice is not possible because its calculation over a le is anundecidable problem. Thus, the actual similarity metric is based on a real lifecompressor which only approximates the Kolmogorov complexity. To use themetric for distance measurement of Bulgarian dialects we rst represent thedialectological data in such a way that the metric is applicable. We proposetwo such representations which are compared to a baseline distance betweendialects. Then we conclude the paper with an outline of our future work.On the Vertex Separation of Cactus Graphs
http://hdl.handle.net/10525/333
Title: On the Vertex Separation of Cactus Graphs<br/><br/>Authors: Markov, Minko<br/><br/>Abstract: This paper is part of a work in progress whose goal is to construct a fast, practical algorithm for the vertex separation (VS) of cactus graphs. We prove a \main theorem for cacti", a necessary and sufficient condition for the VS of a cactus graph being k. Further, we investigate the ensuing ramifications that prevent the construction of an algorithm based on that theorem only.Mixed Theories
http://hdl.handle.net/10525/332
Title: Mixed Theories<br/><br/>Authors: Radev, Slavian<br/><br/>Abstract: In the present paper we investigate the life cycles of formalizedtheories that appear in decision making instruments and science. In fewwords mixed theories are build in the following steps: Initially a smallcollection of facts is the kernel of the theory. To express these facts wemake a special formalized language. When the collection grows we addsome inference rules and thus some axioms to compress the knowledge. Thenext step is to generalize these rules to all expressions in the formalizedlanguage. For these rules we introduce some conclusion procedure. In sucha way we make small theories for restricted fields of the knowledge. The mostimportant procedure is the mixing of these partial knowledge systems. Inthat step we glue the theories together and eliminate the contradictions. Thelast operation is the most complicated one and some simplifying proceduresare proposed.Experiments with two Approaches for Tracking Drifting Concepts
http://hdl.handle.net/10525/331
Title: Experiments with two Approaches for Tracking Drifting Concepts<br/><br/>Authors: Koychev, Ivan<br/><br/>Abstract: This paper addresses the task of learning classifiers from streams of labelled data. In this case we can face the problem that the underlying concepts can change over time. The paper studies two mechanisms developed for dealing with changing concepts. Both are based on the time window idea. The first one forgets gradually, by assigning to the examples weight that gradually decreases over time. The second one uses a statistical test to detect changes in concept and then optimizes the size of the time window, aiming to maximise the classification accuracy on the new examples. Both methods are general in nature and can be used with any learning algorithm. The objectives of the conducted experiments were to compare the mechanisms and explore whether they can be combined to achieve a synergetic e ect. Results from experiments with three basic learning algorithms (kNN, ID3 and NBC) using four datasets are reported and discussed.On Multiple Deletion Codes
http://hdl.handle.net/10525/330
Title: On Multiple Deletion Codes<br/><br/>Authors: Landjev, Ivan; Haralambiev, Kristiyan<br/><br/>Abstract: In 1965 Levenshtein introduced the deletion correcting codes and found an asymptotically optimal family of 1-deletion correcting codes. During the years there has been a little or no research on t-deletion correcting codes for larger values of t. In this paper, we consider the problem of finding the maximal cardinality L2(n;t) of a binary t-deletion correcting code of length n. We construct an infinite family of binary t-deletion correcting codes. By computer search, we construct t-deletion codes for t = 2;3;4;5 with lengths n ≤ 30. Some of these codes improve on earlier results by Hirschberg-Fereira and Swart-Fereira. Finally, we prove a recursive upper bound on L2(n;t) which is asymptotically worse than the best known bounds, but gives better estimates for small values of n.Apollo 13 Risk Assessment Revisited
http://hdl.handle.net/10525/329
Title: Apollo 13 Risk Assessment Revisited<br/><br/>Authors: Bukovics, István<br/><br/>Abstract: Fault tree methodology is the most widespread risk assessmenttool by which one is able to predict - in principle - the outcome of an eventwhenever it is reduced to simpler ones by the logic operations conjunctionand disjunction according to the basics of Boolean algebra. The objectof this work is to present an algorithm by which, using the correspondingcomputer code, one is able to predict - in practice - the outcome of an eventwhenever its fault tree is given in the usual form.Comparative Analysis: A Feasible Software Engineering Method
http://hdl.handle.net/10525/328
Title: Comparative Analysis: A Feasible Software Engineering Method<br/><br/>Authors: Maneva, Nelly<br/><br/>Abstract: The reasonable choice is a critical success factor for decision-making in the field of software engineering (SE). A case-driven comparativeanalysis has been introduced and a procedure for its systematic applicationhas been suggested. The paper describes how the proposed method can bebuilt in a general framework for SE activities. Some examples of experimentalversions of the framework are briey presented.