Abstract
Many well-known machine learning algorithms have been applied to the task of time-series classification, including decision trees, neural networks, support vector machines and others. However, it was shown that the simple 1-nearest neighbor (1NN) classifier, coupled with an elastic distance measure like Dynamic Time Warping (DTW), often produces better results than more complex classifiers on time-series data, including k-nearest neighbor (kNN) for values of \(k>1\). In this article, we revisit the kNN classifier on time-series data by considering ten classic distance-based vote weighting schemes in the context of Euclidean distance, as well as four commonly used elastic distance measures: DTW, Longest Common Subsequence, Edit Distance with Real Penalty and Edit Distance on Real sequence. Through experiments on the complete collection of UCR time-series datasets, we confirm the view that the 1NN classifier is very hard to beat. Overall, for all considered distance measures, we found that variants of the Dudani weighting scheme produced the best results.
Similar content being viewed by others
References
Agrawal R, Faloutsos C, Swami A (1993) Efficient similarity search in sequence databases. In: Lomet David B (ed) Proceedings of the 4th international conference on foundations of data organization and algorithms (FODO’93). Springer, Berlin Heidelberg, pp 69–84
Bache K, Lichman M (2013) UCI machine learning repository. University of California, School of Information and Computer Science, Irvine, CA
Berndt D, Clifford J (1994) Using dynamic time warping to find patterns in time series. In: Usama M, Fayyad RU (ed) Knowledge discovery in databases: papers from the 1994 AAAI workshop. AAAI Press, Seattle, Washington, pp 359–370
Bouckaert RR, Frank E (2004) Evaluating the replicability of significance tests for comparing learning algorithms. In: Dai H, Srikant R, Zhang C (eds) Advances in knowledge discovery and data mining. Springer, Berlin, Heidelberg, pp 3–12
Brockwell PJ, Davis RA (2002) Introduction to time series and forecasting. Springer, New York
Chen L, Ng R (2004) On the marriage of lp-norms and edit distance. In: Nascimento MA, Özsu MT, Kossmann D, et al. (eds) Proceedings of the thirtieth international conference on very large data bases, Toronto, Canada, August 31–September 3, 2004. Morgan Kaufmann, pp 792–803
Chen L, Özsu MT, Oria V (2005) Robust and fast similarity search for moving object trajectories. In: Proceedings of the 2005 ACM SIGMOD international conference on management of data—SIGMOD’05. ACM Press, New York, New York, USA, pp 491–502
Cover T, Hart P (1967) Nearest neighbor pattern classification. IEEE Trans Inf Theory 13:21–27
Das G, Gunopulos D (2003) Time series similarity and indexing. In: Ye N (ed) The handbook of data mining. Lawrence Erlbaum Associates, Mahwah, pp 279–304
Ding H, Trajcevski G, Scheuermann P et al (2008) Querying and mining of time series data: experimental comparison of representations and distance measures. In: Proceedings of the VLDB endowment, vol 1. pp 1542–1552
Dudani SA (1976) The distance-weighted k-nearest-neighbor rule. IEEE Trans Syst Man Cybern SMC 6:325–327
Esling P, Agon C (2012) Time-series data mining. ACM Comput Surv 45:12:1–12:34
Faloutsos C, Ranganathan M, Manolopoulos Y (1994) Fast subsequence matching in time-series databases. In: Proceedings of ACM SIGMOD record, vol 23. pp 419–429
Fix E, Hodges JL (1989) Discriminatory analysis. Nonparametric discrimination: consistency properties. Int Stat Rev 57:238–247
García S, Fernández A, Luengo J, Herrera F (2009) A study of statistical techniques and performance measures for genetics-based machine learning: accuracy and interpretability. Soft Comput 13:959–977
García S, Herrera F (2008) An extension on “statistical comparisons of classifiers over multiple data sets” for all pairwise comparisons. J Mach Learn Res 9:2677–2694
Goldin DQ, Kanellakis PC (1995) On similarity queries for time-series data: Constraint specification and implementation. In: Montanari U, Rossi F (eds) Proceedings of principles and practice of constraint programming—CP’95. Springer, Berlin Heidelberg, pp 137–153
Górecki T, Łuczak M (2013) Using derivatives in time series classification. Data Min Knowl Discov 26:310–331
Gou J, Du L, Zhang Y, Xiong T (2012) A new distance-weighted k-nearest neighbor classifier. J Inf Comput Sci 9:1429–1436
Gou J, Xiong T, Kuang Y (2011) A novel weighted voting for k-nearest neighbor rule. J Comput 6:833–840
Han J, Kamber M (2006) Data mining: concepts and techniques, 2nd edn. Morgan Kaufmann Publishers Inc., San Francisco
Hand DJ, Mannila H, Smyth P (2001) Principles of data mining. MIT Press, Cambridge
Itakura F (1975) Minimum prediction residual principle applied to speech recognition. IEEE Trans Acoust Speech Signal Process 23:67–72
Jeong Y-S, Jeong MK, Omitaomu OA (2011) Weighted dynamic time warping for time series classification. Pattern Recogn 44:2231–2240
Keogh E, Ratanamahatana CA (2005) Exact indexing of dynamic time warping. Knowl Inf Syst 7:358–386
Keogh E, Zhu Q, Hu B, et al (2011) The UCR time series classification/clustering homepage. www.cs.ucr.edu/~eamonn/time_series_data/
Kotsiantis SB (2007) Supervised machine learning: a review of classification techniques. In: Proceedings of the 2007 conference on emerging artificial intelligence applications in computer engineering: real word AI systems with applications in eHealth, HCI. Information retrieval and pervasive technologies. IOS Press, Amsterdam, pp 3–24
Kurbalija V, Ivanović M, Budimac Z (2009) Case-based curve behaviour prediction. Softw Pract Exp 39:81–103
Kurbalija V, Ivanović M, von Bernstorff C et al (2014a) Matching observed with empirical reality—what you see is what you get? Fundam Inform 129:133–147
Kurbalija V, Radovanović M, Geler Z, Ivanović M (2010) A framework for time-series analysis. In: Dicheva D, Dochev D (eds) Artificial intelligence: methodology, systems, and applications SE-5. Springer, Berlin, Heidelberg, pp 42–51
Kurbalija V, Radovanović M, Geler Z, Ivanović M (2011) The Influence of Global Constraints on DTW and LCS Similarity Measures for Time-Series Databases. In: Dicheva D, Markov Z, Stefanova E (eds) Third international conference on software, services and semantic technologies S3T 2011 SE-10. Springer, Berlin, Heidelberg, pp 67–74
Kurbalija V, Radovanović M, Geler Z, Ivanović M (2014b) The influence of global constraints on similarity measures for time-series databases. Knowl Based Syst 56:49–67
Kurbalija V, Radovanović M, Ivanović M et al (2014c) Time-series analysis in the medical domain: a study of Tacrolimus administration and influence on kidney graft function. Comput Biol Med 50:19–31
Kurbalija V, von Bernstorff C, Burkhard H-D et al (2012) Time-series mining in a psychological domain. In: Proceedings of the fifth Balkan conference in informatics on—BCI ’12. ACM Press, New York, New York, USA, pp 58–63
Larose DT (2005) Discovering knowledge in data. Wiley, Hoboken
Laxman S, Sastry PS (2006) A survey of temporal data mining. Sadhana 31:173–198
Macleod J, Luk A, Titterington D (1987) A re-examination of the distance-weighted k-nearest neighbor classification rule. IEEE Trans Syst Man Cybern 17:689–696
Marteau P-F (2009) Time warp edit distance with stiffness adjustment for time series matching. IEEE Trans Pattern Anal Mach Intell 31:306–318
Mitchell TM (1997) Mach Learn. McGraw-Hill Inc, New York
Mitrović D, Geler Z, Ivanović M (2012) Distributed distance matrix generator based on agents. In: Proceedings of the 2nd international conference on web intelligence, mining and semantics—WIMS’12. ACM Press, New York, New York, USA, pp 40:1–40:6
Mitrovic D, Ivanović M, Geler Z (2014) Agent-based distributed computing for dynamic networks. Inf Technol Control 43:88–97
Morse MD, Patel JM (2007) An efficient and accurate method for evaluating time series similarity. In: Proceedings of the 2007 ACM SIGMOD international conference on management of data—SIGMOD’07. ACM Press, New York, New York, USA, pp 569–580
Nanopoulos A, Alcock R, Manolopoulos Y (2001) Feature-based classification of time-series data. Int J Comput Res 10:49–61
Pao T-L, Chen Y-T, Yeh J-H et al (2007) A comparative study of different weighting schemes on KNN-based emotion recognition in Mandarin speech. In: Huang D-S, Heutte L, Loog M (eds) Advanced intelligent computing theories and applications. With aspects of theoretical and methodological issues. Springer, Berlin, Heidelberg, pp 997–1005
Pao T-L, Chen Y-T, Yeh J-H, Chang Y-H (2005) Emotion recognition and evaluation of Mandarin speech using weighted D-KNN classification. In: Proceedings of the 17th conference on computational linguistics and speech processing, ROCLING 2005, Taiwan, ROC, 2005. Association for Computational Linguistics and Chinese Language Processing (ACLCLP), Taiwan, pp 203–212
Pavlovic V, Frey BJ, Huang TS (1999) Time-series classification using mixed-state dynamic Bayesian networks. In: Proceedings of 1999 IEEE computer society conference on computer vision and pattern recognition (Cat. No PR00149). IEEE Computer Society, pp 609–615
Radovanović M, Nanopoulos A, Ivanović M (2010) Time-series classification in many intrinsic dimensions. In: Proceedings of the 2010 SIAM international conference on data mining. Society for Industrial and Applied Mathematics, Philadelphia, PA, pp 677–688
Radovanović M, Nanopoulos A, Ivanović M (2010b) Hubs in space: popular nearest neighbors in high-dimensional data. J Mach Learn Res 11:2487–2531
Ralanamahatana CA, Lin J, Gunopulos D et al (2005) Mining time series data. In: Maimon O, Rokach L (eds) Data mining and knowledge discovery handbook. Springer, New York, pp 1069–1103
Ratanamahatana CA, Keogh E (2005) Three myths about dynamic time warping data mining. In: Proceedings of the 2005 SIAM international conference on data mining. Society for Industrial and Applied Mathematics, Philadelphia, PA, pp 506–510
Rodríguez JJ, Alonso CJ (2004) Interval and dynamic time warping-based decision trees. In: Proceedings of the 2004 ACM symposium on applied computing. ACM, New York, NY, USA, pp 548–552
Rodríguez JJ, Alonso CJ, Boström H (2000) Learning first order logic time series classifiers: rules and boosting. In: Zighed D, Komorowski J, Żytkow J (eds) Principles of data mining and knowledge discovery SE-29. Springer, Berlin, Heidelberg, pp 299–308
Sakoe H, Chiba S (1978) Dynamic programming algorithm optimization for spoken word recognition. IEEE Trans Acoust Speech Signal Process 26:43–49
Shi T, Wang P, Wang J-S, Yue S (2012) Application of grid-based k-means clustering algorithm for optimal image processing. Comput Sci Inf Syst 9:1679–1696
Skopal T, Bustos B (2011) On nonmetric similarity search problems in complex domains. ACM Comput Surv 43:34:1–34:50
Stojanović R, Knežević S, Karadaglić D, Devedžić G (2013) Optimization and implementation of the wavelet based algorithms for embedded biomedical signal processing. Comput Sci Inf Syst 10:503–523
Takigawa Y, Hott S, Kiyasu S, Miyahara S (2005) Pattern classification using weighted average patterns of categorical k-nearest neighbors. In: Proceedings of the 1th international workshop on camera-based document analysis and recognition. pp 111–118
Tomašev N, Mladenić D (2012) Nearest neighbor voting in high dimensional data: learning from past occurrences. Comput Sci Inf Syst 9:691–712
Vlachos M, Kollios G, Gunopulos D (2002) Discovering similar multidimensional trajectories. In: Proceedings 18th international conference on data engineering. IEEE Computer Society, pp 673–684
Wu X, Kumar V, Ross Quinlan J et al (2008) Top 10 algorithms in data mining. Knowl Inf Syst 14:1–37
Wu Y, Chang EY (2004) Distance-function design and fusion for sequence data. In: Proceedings of the thirteenth ACM conference on information and knowledge management—CIKM’04. ACM Press, New York, New York, USA, pp 324–333
Wu Y-L, Agrawal D, El Abbadi A (2000) A comparison of DFT and DWT based similarity search in time-series databases. In: Proceedings of the ninth international conference on Information and knowledge management—CIKM ’00. ACM Press, New York, New York, USA, pp 488–495
Xi X, Keogh E, Shelton C et al (2006) Fast time series classification using numerosity reduction. In: Proceedings of the 23rd international conference on machine learning—ICML’06. ACM Press, New York, New York, USA, pp 1033–1040
Ye L, Keogh E (2009) Time series shapelets. In: Proceedings of the 15th ACM SIGKDD international conference on knowledge discovery and data mining—KDD’09. ACM Press, New York, New York, USA, pp 947–956
Zavrel J (1997) An empirical re-examination of weighted voting for k-NN. In: Proceedings of the 7th Belgian–Dutch conference on machine learning. pp 139–148
Zhang H, Ho TB, Lin MS (2004) A non-parametric wavelet feature extractor for time series classification. In: Dai H, Srikant R, Zhang C (eds) Advances in knowledge discovery and data mining SE-71. Springer, Berlin, Heidelberg, pp 595–603
Acknowledgments
The authors would like to thank Eamonn Keogh for collecting and making available the UCR time-series datasets, as well as everyone who contributed data to the collection, without whom the presented work would not have been possible. V. Kurbalija, M. Radovanović and M. Ivanović thank the Serbian Ministry of Education, Science and Technological Development for support through Project No. OI174023, “Intelligent Techniques and their Integration into Wide-Spectrum Decision Support.”
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Tables 20, 21, 22, 23 and 24 contain the classification errors and the values of parameter k obtained for the analyzed similarity measures (Euclidean distance, DTW, LCS, ERP and EDR). Due to lack of space, the values reported in the tables in the Appendix are shown rounded to three decimal places.
Rights and permissions
About this article
Cite this article
Geler, Z., Kurbalija, V., Radovanović, M. et al. Comparison of different weighting schemes for the kNN classifier on time-series data. Knowl Inf Syst 48, 331–378 (2016). https://doi.org/10.1007/s10115-015-0881-0
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10115-015-0881-0