The IRMA Community
Newsletters
Research IRM
Click a keyword to search titles using our InfoSci-OnDemand powered search:
|
Building Ensembles Using Decision Tree Metrics Based Meta-Trees
Abstract
Ensembles of classifiers have become one of the most popular techniques in machine learning research. The problem of selecting the most appropriate classifier for classification that is known from machine learning is also an important issue in the ensemble building methods. This paper tries to find a method that would select the best performing ensemble building method based on dataset characteristics instead of the most appropriate classifier. The proposed approach captures the characterization of the dataset from the metrics of the tree induced from the dataset. On 15 benchmark datasets, the proposed meta-tree based method discovered some strong and simple rules that could be used in future research in the field of basic ensemble building method selection.
|
|