CHI-BD: A fuzzy rule-based classification system for Big Data classification problems

The previous Fuzzy Rule-Based Classification Systems (FRBCSs) for Big Data problems consist in concurrently learning multiple Chi et al. FRBCSs whose rule bases are then aggregated. The problem of this approach is that different models are obtained when varying the configuration of the cluster, beco...

Full description

Saved in:
Bibliographic Details
Published in:Fuzzy sets and systems 2018-10, Vol.348, p.75-101
Main Authors: Elkano, Mikel, Galar, Mikel, Sanz, Jose, Bustince, Humberto
Format: Article
Language:English
Subjects:
Big Data
Fuzzy Rule-Based Classification Systems
Hadoop
Imbalanced datasets
MapReduce
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The previous Fuzzy Rule-Based Classification Systems (FRBCSs) for Big Data problems consist in concurrently learning multiple Chi et al. FRBCSs whose rule bases are then aggregated. The problem of this approach is that different models are obtained when varying the configuration of the cluster, becoming less accurate as more computing nodes are added. Our aim with this work is to design a new FRBCS for Big Data classification problems (CHI-BD) which is able to provide exactly the same model as the one that would be obtained by the original Chi et al. algorithm if it could be executed with this quantity of data. In order to do so, we take advantage of the suitability of the Chi et al. algorithm for the MapReduce paradigm, solving the problems of the previous approach, which lead us to obtain the same model (i.e., classification accuracy) regardless of the number of computing nodes considered.
ISSN:0165-0114
1872-6801
DOI:10.1016/j.fss.2017.07.003