Research Repository

See what's under the surface

A scalable expressive ensemble learning using random prism: a MapReduce approach.

Stahl, Frederic; May, David; Mills, Hugo; Bramer, Max; Gaber, Mohamed Medhat

Authors

Frederic Stahl

David May

Hugo Mills

Max Bramer

Mohamed Medhat Gaber



Contributors

Abdelkader Hameurlain
Editor

Josef Küng
Editor

Roland Wagner
Editor

Sherif Sakr
Editor

Lizhe Wang
Editor

Albert Zomaya
Editor

Abstract

The induction of classification rules from previously unseen examples is one of the most important data mining tasks in science as well as commercial applications. In order to reduce the infl uence of noise in the data, ensemble learners are often applied. However, most ensemble learners are based on decision tree classifiers which are affected by noise. The Random Prism classifier has recently been proposed as an alterna tive to the popular Random Forests classifier, which is based on decision trees. Random Prism is based on the Prism family of algorithms, which is more robust to noise. However, like most ensemble classification ap proaches, Random Prism also does not scale well on large training data. This paper presents a thorough discussion of Random Prism and a re cently proposed parallel version of it called Parallel Random Prism. Par allel Random Prism is based on the MapReduce programming paradigm. The paper provides, for the first time, novel theoretical analysis of the proposed technique and in-depth experimental study that show that Par allel Random Prism scales well on a large number of training examples, a large number of data features and a large number of processors. Ex pressiveness of decision rules that our technique produces makes it a natural choice for Big Data applications where informed decision mak ing increases the user's trust in the system.

Publication Date Mar 18, 2015
Print ISSN 0302-9743
Publisher Springer (part of Springer Nature)
Pages 90-107
Series Title Lecture notes in computer science
Series Number 9070
Series ISSN 0302-9743
Book Title Transactions on large-scale data- and knowledge-centred systems XX: special issue on advanced techniques for big data management.
Institution Citation STAHL, F., MAY, D., MILLS, H., BRAMER, M. and GABER, M.M. 2015. A scalable expressive ensemble learning using random prism: a MapReduce approach. In Hameurlain, A., K√ľng, J., Wagner, R., Sakr, S., Wang, L. and Zomaya, A. (eds.) Transactions on large-scale data- and knowledge-centred systems XX: special issue on advanced techniques for big data management. Lecture notes in computer science, 9070. Berlin: Springer [online], pages 90-107. Available from: https://doi.org/10.1007/978-3-662-46703-9_4
DOI https://doi.org/10.1007/978-3-662-46703-9_4
Keywords Random Forest; Base Classifier; Computing Node; Training Instance; Ensemble Classifier

Files




Downloadable Citations