Fcnn-Mr: A Parallel Instance Selection Method Based On Fast Condensed Nearest Neighbor Rule
Lists
Only you see this
Abstract
Instance selection (IS) technique is used to reduce<br> the data size to improve the performance of data mining methods.<br> Recently, to process very large data set, several proposed methods<br> divide the training set into some disjoint subsets and apply IS<br> algorithms independently to each subset. In this paper, we analyze<br> the limitation of these methods and give our viewpoint about how to<br> divide and conquer in IS procedure. Then, based on fast condensed<br> nearest neighbor (FCNN) rule, we propose a large data sets instance<br> selection method with MapReduce framework. Besides ensuring the<br> prediction accuracy and reduction rate, it has two desirable properties:<br> First, it reduces the work load in the aggregation node; Second<br> and most important, it produces the same result with the sequential<br> version, which other parallel methods cannot achieve. We evaluate the<br> performance of FCNN-MR on one small data set and two large data<br> sets. The experimental results show that it is effective and practical.
BibTeX
@inproceedings{si2017fcnn,
author = {Lu Si and Jie Yu and Shasha Li and Jun Ma and Lei Luo and Qingbo Wu and Yongqi Ma and Zhengji Liu},
title = {{Fcnn-Mr: A Parallel Instance Selection Method Based On Fast Condensed Nearest Neighbor Rule}},
booktitle = {Zenodo},
year = {2017},
doi = {10.5281/zenodo.1131427},
}Loading related content...