{"references": ["S. Garc\u0142a, J. Luengo, and F. Herrera, \"Data preprocessing in data\nmining,\" Computer Science, vol. 72, 2015.", "B. J. Han, \"Data mining. concepts and techniques. 3rd ed,\" Data Mining\nConcepts Models Methods & Algorithms Second Edition, vol. 5, no. 4,\npp. 1 \u2013 18, 2000.", "T. Cover and P. Hart, \"Nearest neighbor pattern classification,\" IEEE\nTransactions on Information Theory, vol. 13, no. 1, pp. 21\u201327, 1967.", "D. R. Wilson and T. R. Martinez, \"Reduction techniques for\ninstance-based learning algorithms,\" Machine Learning, vol. 38, no. 3,\npp. 257\u2013286, 2000.", "M. Kudo and J. Sklansky, \"Comparison of algorithms that select features\nfor pattern classifiers,\" Pattern Recognition, vol. 33, no. 1, pp. 25\u201341,\n2000.", "H. Liu and H. Motoda, \"Feature extraction construction and selection:\nA data mining perspective,\" Springer International, vol. 94, no. 448, p.\n014004, 1999.", "I. Triguero, J. Derrac, S. Garcia, and F. Herrera, \"A taxonomy\nand experimental study on prototype generation for nearest neighbor\nclassification,\" Systems Man & Cybernetics Part C Applications &\nReviews IEEE Transactions on, vol. 42, no. 1, pp. 86\u2013100, 2012.", "J. Hamidzadeh, R. Monsefi, and H. S. Yazdi, \"Irahc: Instance reduction\nalgorithm using hyperrectangle,\" Pattern Recognition, vol. 48, no. 5, pp.\n1878\u20131889, 2015.", "Haro-Garc, A. Aida, Garc, and N. A-Pedrajas, \"A divide-and-conquer\nrecursive approach for scaling up instance selection algorithms,\" Data\nMining and Knowledge Discovery, vol. 18, no. 3, pp. 392\u2013418, 2009.\n[10] I. Triguero, D. Peralta, J. Bacardit, S. Garc\u0142a, and F. Herrera, \"Mrpr: A\nmapreduce solution for prototype reduction in big data classification,\"\nNeurocomputing, vol. 150, no. 150, p. 331C345, 2015.\n[11] J. Zhai, X. Wang, and X. Pang, \"Voting-based instance selection\nfrom large data sets with mapreduce and random weight networks,\"\nInformation Sciences, vol. 367, pp. 1066\u20131077, 2016.\n[12] H. Liu and H. Motoda, \"On issues of instance selection.\" Data Mining\nand Knowledge Discovery, vol. 6, no. 2, pp. 115\u2013130, 2002.\n[13] J. R. Cano, F. Herrera, and M. Lozano, \"Stratification for scaling up\nevolutionary prototype selection,\" Pattern Recognition Letters, vol. 26,\nno. 7, pp. 953\u2013963, 2005.\n[14] J. Dean and S. Ghemawat, \"Mapreduce: Simplified data processing\non large clusters.\" in Conference on Symposium on Opearting Systems\nDesign & Implementation, 2004, pp. 107\u2013113.\n[15] F. Angiulli, \"Fast condensed nearest neighbor rule,\" in International\nConference, 2005, pp. 25\u201332.\n[16] Angiulli, \"Fast nearest neighbor condensation for large data sets\nclassification,\" IEEE Transactions on Knowledge & Data Engineering,\nvol. 19, no. 11, pp. 1450\u20131464, 2007.\n[17] B. P. E. Hart, \"The condensed nearest neighbor rule,\" in IEEE Trans.\nInformation Theory, 1968.\n[18] C. H. Chou, B. H. Kuo, and F. Chang, \"The generalized condensed\nnearest neighbor rule as a data reduction method,\" vol. 2, pp. 556\u2013559,\n2006.\n[19] G. W. Gates, \"The reduced nearest neighbor rule,\" IEEE Transactions\non Information Theory, vol. 18, no. 3, pp. 431 \u2013 433, 1972.\n[20] D. L. Wilson, \"Asymptotic properties of nearest neighbor rules using\nedited data,\" IEEE Transactions on Systems Man & Cybernetics, vol. 2,\nno. 3, pp. 408\u2013421, 1972.\n[21] W. C. Lin, C. F. Tsai, S. W. Ke, C. W. Hung, and W. Eberle, \"Learning\nto detect representative data for large scale instance selection,\" Journal\nof Systems & Software, vol. 106, no. C, pp. 1\u20138, 2015.\n[22] A. Onan, \"A fuzzy-rough nearest neighbor classifier combined with\nconsistency-based subset evaluation and instance selection for automated\ndiagnosis of breast cancer,\" Expert Systems with Applications, vol. 42,\nno. 20, pp. 6844\u20136852, 2015.\n[23] J. A. Olvera-Lpez, J. A. Carrasco-Ochoa, and J. F. Mart\u0142nez-Trinidad,\n\"A new fast prototype selection method based on clustering,\" Pattern\nAnalysis and Applications, vol. 13, no. 2, pp. 131\u2013141, 2010.\n[24] J. R. Cano, F. Herrera, and M. Lozano, \"Using evolutionary algorithms\nas instance selection for data reduction in kdd: an experimental study,\"\nIEEE Transactions on Evolutionary Computation, vol. 7, no. 6, pp.\n561\u2013575, 2004.\n[25] D. Borthakur, \"The hadoop distributed file system: Architecture and\ndesign,\" Hadoop Project Website, vol. 11, no. 11, pp. 1 \u2013 10, 2007.\n[26] D. B. Skalak, \"Prototype and feature selection by sampling and random\nmutation hill climbing algorithms,\" Machine Learning Proceedings, pp.\n293\u2013301, 1994.\n[27] V. S. Devi and M. N. Murty, \"An incremental prototype set building\ntechnique,\" Pattern Recognition, vol. 35, no. 2, pp. 505\u2013513, 2002."]}

Instance selection (IS) technique is used to reduce the data size to improve the performance of data mining methods. Recently, to process very large data set, several proposed methods divide the training set into some disjoint subsets and apply IS algorithms independently to each subset. In this paper, we analyze the limitation of these methods and give our viewpoint about how to divide and conquer in IS procedure. Then, based on fast condensed nearest neighbor (FCNN) rule, we propose a large data sets instance selection method with MapReduce framework. Besides ensuring the prediction accuracy and reduction rate, it has two desirable properties: First, it reduces the work load in the aggregation node; Second and most important, it produces the same result with the sequential version, which other parallel methods cannot achieve. We evaluate the performance of FCNN-MR on one small data set and two large data sets. The experimental results show that it is effective and practical.