We give an unified convergence analysis of ensemble learning methods including e.g. AdaBoost, Logistic Regression and the Least-SquareBoost algorithm for regression. These methods...
We present compelling evidence that the World Wide Web is a domain in which applications can benefit from using first-order learning methods, since the graph structure inherent in ...
This paper proposes an extension of Sumida and Torisawa's method of acquiring hyponymy relations from hierachical layouts in Wikipedia (Sumida and Torisawa, 2008). We extract...
We have explored a novel method to find textual relations in electronic documents using genetic programming and semantic networks. This can be used for enhancing information retri...
Multi-relational data mining has become popular due to the limitations of propositional problem definition in structured domains and the tendency of storing data in relational dat...
Yusuf Kavurucu, Pinar Senkul, Ismail Hakki Toroslu