The Experts below are selected from a list of 291 Experts worldwide ranked by ideXlab platform
Yves Lepage - One of the best experts on this subject based on the ideXlab platform.
-
An Investigation of the Sampling-Based Alignment Method and Its Contributions
International Journal of Artificial Intelligence & Applications, 2013Co-Authors: Juan Luo, Yves LepageAbstract:By investigating the distribution of phrase pairs in phrase translation tables, the work in this paper describes an approach to increase the number of n-gram alignments in phrase translation tables output by a sampling-based alignment method. This approach consists in enforcing the alignment of n-grams in distinct translation Subtables so as to increase the number of n-grams. Standard normal distribution is used to allot alignment time among translation Subtables, which results in adjustment of the distribution of ngrams. This leads to better evaluation results on statistical machine translation tasks than the original sampling-based alignment approach. Furthermore, the translation quality obtained by merging phrase translation tables computed from the sampling-based alignment method and from MGIZA++ is examined.
Beata Zielosko - One of the best experts on this subject based on the ideXlab platform.
-
Sequential Optimization of \(\gamma \)-Decision Rules Relative to Length, Coverage and Number of Misclassifications
Transactions on Rough Sets XIX, 2015Co-Authors: Beata ZieloskoAbstract:The paper is devoted to the study of an extension of dynamic programming approach which allows sequential optimization of approximate decision rules relative to length, coverage and number of misclassifications. Presented algorithm constructs a directed acyclic graph \({\varDelta }_{\gamma }(T)\) which nodes are Subtables of the decision table T. Based on the graph \({\varDelta }_{\gamma }(T)\) we can describe all irredundant \(\gamma \)-decision rules with the minimum length, after that among these rules describe all rules with the maximum coverage, and among such rules describe all rules with the minimum number of misclassifications. We can also change the set of cost functions and order of optimization. Sequential optimization can be considered as a tool that helps to construct simpler rules for understanding and interpreting by experts.
-
Greedy Algorithm for Construction of Decision Trees for Tables with Many-Valued Decisions
2014Co-Authors: Mohammad Azad, Mikhail Moshkov, Igor Chikalov, Beata ZieloskoAbstract:Abstract. In the paper, we study a greedy algorithm for construction of approximate decision trees. This algorithm is applicable to decision tables with many-valued decisions where each row is labeled with a set of decisions. For a given row, we should find a decision from the set attached to this row. We use an uncertainty measure which is the number of boundary Subtables. We present also experimental results for data sets from UCI Machine Learning Repository for proposed approach and approach based on generalized decision
-
Innovations in Intelligent Machines (4) - Optimization of Decision Rules Based on Dynamic Programming Approach
Studies in Computational Intelligence, 2013Co-Authors: Beata Zielosko, Igor Chikalov, Mikhail Moshkov, Talha AminAbstract:This chapter is devoted to the study of an extension of dynamic programming approach which allows optimization of approximate decision rules relative to the length and coverage. We introduce an uncertainty measure that is the difference between number of rows in a given decision table and the number of rows labeled with the most common decision for this table divided by the number of rows in the decision table. We fix a threshold γ, such that 0 ≤ γ < 1, and study so-called γ-decision rules (approximate decision rules) that localize rows in Subtables which uncertainty is at most γ. Presented algorithm constructs a directed acyclic graph Δγ T which nodes are Subtables of the decision table T given by pairs “attribute = value”. The algorithm finishes the partitioning of a subtable when its uncertainty is at most γ. The chapter contains also results of experiments with decision tables from UCI Machine Learning Repository.
-
dynamic programming approach to optimization of approximate decision rules
Information Sciences, 2013Co-Authors: Talha Amin, Mikhail Moshkov, Igor Chikalov, Beata ZieloskoAbstract:This paper is devoted to the study of an extension of dynamic programming approach which allows sequential optimization of approximate decision rules relative to the length and coverage. We introduce an uncertainty measure R(T) which is the number of unordered pairs of rows with different decisions in the decision table T. For a nonnegative real number @b, we consider @b-decision rules that localize rows in Subtables of T with uncertainty at most @b. Our algorithm constructs a directed acyclic graph @D"@b(T) which nodes are Subtables of the decision table T given by systems of equations of the kind ''attribute=value''. This algorithm finishes the partitioning of a subtable when its uncertainty is at most @b. The graph @D"@b(T) allows us to describe the whole set of so-called irredundant @b-decision rules. We can describe all irredundant @b-decision rules with minimum length, and after that among these rules describe all rules with maximum coverage. We can also change the order of optimization. The consideration of irredundant rules only does not change the results of optimization. This paper contains also results of experiments with decision tables from UCI Machine Learning Repository.
-
KES (Selected Papers) - Optimization of approximate decision rules relative to number of misclassifications: comparison of greedy and dynamic programming approaches
Lecture Notes in Computer Science, 2013Co-Authors: Talha Amin, Mikhail Moshkov, Igor Chikalov, Beata ZieloskoAbstract:In the paper, we present a comparison of dynamic programming and greedy approaches for construction and optimization of approximate decision rules relative to the number of misclassifications. We use an uncertainty measure that is a difference between the number of rows in a decision table T and the number of rows with the most common decision for T. For a nonnegative real number γ, we consider γ-decision rules that localize rows in Subtables of T with uncertainty at most γ. Experimental results with decision tables from the UCI Machine Learning Repository are also presented.
Juan Luo - One of the best experts on this subject based on the ideXlab platform.
-
An Investigation of the Sampling-Based Alignment Method and Its Contributions
International Journal of Artificial Intelligence & Applications, 2013Co-Authors: Juan Luo, Yves LepageAbstract:By investigating the distribution of phrase pairs in phrase translation tables, the work in this paper describes an approach to increase the number of n-gram alignments in phrase translation tables output by a sampling-based alignment method. This approach consists in enforcing the alignment of n-grams in distinct translation Subtables so as to increase the number of n-grams. Standard normal distribution is used to allot alignment time among translation Subtables, which results in adjustment of the distribution of ngrams. This leads to better evaluation results on statistical machine translation tasks than the original sampling-based alignment approach. Furthermore, the translation quality obtained by merging phrase translation tables computed from the sampling-based alignment method and from MGIZA++ is examined.
Mikhail Moshkov - One of the best experts on this subject based on the ideXlab platform.
-
Some Tools for Decision Tables
Extensions of Dynamic Programming for Combinatorial Optimization and Data Mining, 2018Co-Authors: Hassan Aboueisha, Igor Chikalov, Talha Amin, Shahid Hussain, Mikhail MoshkovAbstract:In this chapter, we describe main notions related to decision tables, consider the structure of Subtables of a given decision table represented as a directed acyclic graph (DAG), and discuss time complexity of algorithms on DAGs. We also consider classes of decision tables over restricted information systems for which these algorithms have polynomial time complexity depending on the number of conditional attributes in the input decision table.
-
Greedy Algorithm for Construction of Decision Trees for Tables with Many-Valued Decisions
2014Co-Authors: Mohammad Azad, Mikhail Moshkov, Igor Chikalov, Beata ZieloskoAbstract:Abstract. In the paper, we study a greedy algorithm for construction of approximate decision trees. This algorithm is applicable to decision tables with many-valued decisions where each row is labeled with a set of decisions. For a given row, we should find a decision from the set attached to this row. We use an uncertainty measure which is the number of boundary Subtables. We present also experimental results for data sets from UCI Machine Learning Repository for proposed approach and approach based on generalized decision
-
Innovations in Intelligent Machines (4) - Optimization of Decision Rules Based on Dynamic Programming Approach
Studies in Computational Intelligence, 2013Co-Authors: Beata Zielosko, Igor Chikalov, Mikhail Moshkov, Talha AminAbstract:This chapter is devoted to the study of an extension of dynamic programming approach which allows optimization of approximate decision rules relative to the length and coverage. We introduce an uncertainty measure that is the difference between number of rows in a given decision table and the number of rows labeled with the most common decision for this table divided by the number of rows in the decision table. We fix a threshold γ, such that 0 ≤ γ < 1, and study so-called γ-decision rules (approximate decision rules) that localize rows in Subtables which uncertainty is at most γ. Presented algorithm constructs a directed acyclic graph Δγ T which nodes are Subtables of the decision table T given by pairs “attribute = value”. The algorithm finishes the partitioning of a subtable when its uncertainty is at most γ. The chapter contains also results of experiments with decision tables from UCI Machine Learning Repository.
-
dynamic programming approach to optimization of approximate decision rules
Information Sciences, 2013Co-Authors: Talha Amin, Mikhail Moshkov, Igor Chikalov, Beata ZieloskoAbstract:This paper is devoted to the study of an extension of dynamic programming approach which allows sequential optimization of approximate decision rules relative to the length and coverage. We introduce an uncertainty measure R(T) which is the number of unordered pairs of rows with different decisions in the decision table T. For a nonnegative real number @b, we consider @b-decision rules that localize rows in Subtables of T with uncertainty at most @b. Our algorithm constructs a directed acyclic graph @D"@b(T) which nodes are Subtables of the decision table T given by systems of equations of the kind ''attribute=value''. This algorithm finishes the partitioning of a subtable when its uncertainty is at most @b. The graph @D"@b(T) allows us to describe the whole set of so-called irredundant @b-decision rules. We can describe all irredundant @b-decision rules with minimum length, and after that among these rules describe all rules with maximum coverage. We can also change the order of optimization. The consideration of irredundant rules only does not change the results of optimization. This paper contains also results of experiments with decision tables from UCI Machine Learning Repository.
-
Rough Sets and Intelligent Systems (2) - Relationships for Cost and Uncertainty of Decision Trees
Rough Sets and Intelligent Systems - Professor Zdzisław Pawlak in Memoriam, 2013Co-Authors: Igor Chikalov, Shahid Hussain, Mikhail MoshkovAbstract:This chapter is devoted to the design of new tools for the study of decision trees. These tools are based on dynamic programming approach and need the consideration of Subtables of the initial decision table. So this approach is applicable only to relatively small decision tables. The considered tools allow us to compute: 1 The minimum cost of an approximate decision tree for a given uncertainty value and a cost function. 2 The minimum number of nodes in an exact decision tree whose depth is at most a given value. For the first tool we considered various cost functions such as: depth and average depth of a decision tree and number of nodes (and number of terminal and nonterminal nodes) of a decision tree. The uncertainty of a decision table is equal to the number of unordered pairs of rows with different decisions. The uncertainty of approximate decision tree is equal to the maximum uncertainty of a subtable corresponding to a terminal node of the tree. In addition to the algorithms for such tools we also present experimental results applied to various datasets acquired from UCI ML Repository [4].
D. V. Gokhale - One of the best experts on this subject based on the ideXlab platform.
-
On uniform marginal representation of contingency tables
Statistics & Probability Letters, 1994Co-Authors: Barry C. Arnold, D. V. GokhaleAbstract:Abstract For a two-dimensional probability distribution represented as a contingency table, an algorithm due to Mosteller (1968) “standardizes” the table to have uniform marginals so as to obtain its “uniform marginals representation (UMR)”. In this note, this algorithm is first shown to minimize the Kullback-Leibler information function between distributions with uniform marginals and the given distribution. Next, a characterization of such tables is proved, in terms of (i) their UMRs, (ii) cross-product ratios of each of their 2 × 2 Subtables (iii) the ability to obtain one table from another only by adjustment of their row or column marginals.