Now showing items 11-20 of 47
Univariate decision tree induction using maximum margin classification
(OXFORD UNIV PRESS, 2012-03)
In many pattern recognition applications, first decision trees are used due to their simplicity and easily interpretable nature. In this paper, we propose a new decision tree learning algorithm called univariate margin ...
Omnivariate Rule Induction Using a Novel Pairwise Statistical Test
(IEEE Computer Soc, 2013-09)
Rule learning algorithms, for example, RIPPER, induces univariate rules, that is, a propositional condition in a rule uses only one feature. In this paper, we propose an omnivariate induction of rules where under each ...
Cryptanalysis of Fridrich\'s chaotic image encryption
(WORLD SCIENTIFIC PUBL CO PTE LTD, 2010-05)
We cryptanalyze Fridrich's chaotic image encryption algorithm. We show that the algebraic weaknesses of the algorithm make it vulnerable against chosen-ciphertext attacks. We propose an attack that reveals the secret ...
Incremental construction of rule ensembles using classifiers produced by different class orderings
In this paper, we discuss a novel approach to incrementally construct a rule ensemble. The approach constructs an ensemble from a dynamically generated set of rule classifiers. Each classifier in this set is trained by ...
Shallow parsing in Turkish
In this study, shallow parsing is applied on Turkish sentences. These sentences are used to train and test the per-formances of various learning algorithms with various features specified for shallow parsing in Turkish.
A tree-based approach for English-to-Turkish translation
(Tubitak Scientific & Technical Research Council Turkey, 2019)
In this paper, we present our English-to-Turkish translation methodology, which adopts a tree-based approach. Our approach relies on tree analysis and the application of structural modification rules to get the target side ...
Design and analysis of classifier learning experiments in bioinformatics: survey and case studies
(IEEE COMPUTER SOC, 2012-12)
In many bioinformatics applications, it is important to assess and compare the performances of algorithms trained from data, to be able to draw conclusions unaffected by chance and are therefore significant. Both the design ...
VC-Dimension of Univariate Decision Trees
(IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2015-02)
In this paper, we give and prove the lower bounds of the Vapnik-Chervonenkis (VC)-dimension of the univariate decision tree hypothesis class. The VC-dimension of the univariate decision tree depends on the VC-dimension ...
WikiLeaks on the Middle East: Obscure Diplomacy Networks and Binding Spaces
(ROUTLEDGE JOURNALS, 2014-10-02)
In this paper, we explore the flow of information regarding strategic Middle Eastern countries in the WikiLeaks 'diplomatic cables' by applying data-mining techniques to construct directed networks. The results show that ...
Comparison of Turkish Proposition Banks by Frame Matching
By indicating semantic relations between a predicate and its associated participants in a sentence and identifying the role-bearing constituents, SRL provides an extensive dataset to understand natural languages and to ...