Reliable early classification of time series based on discriminating the classes over time

The goal of early classification of time series is to predict the class value of a sequence early in time, when its full length is not yet available. This problem arises naturally in many contexts where the data is collected over time and the label predictions have to be made as soon as possible. In this work, a method based on probabilistic classifiers is proposed for the problem of early classification of time series. An important feature of this method is that, in its learning stage, it discovers the timestamps in which the prediction accuracy for each class begins to surpass a pre-defined threshold. This threshold is defined as a percentage of the accuracy that would be obtained if the full series were available, and it is defined by the user. The class predictions for new time series will only be made in these timestamps or later. Furthermore, when applying the model to a new time series, a class label will only be provided if the difference between the two largest predicted class probabilities is higher than or equal to a certain threshold, which is calculated in the training step. The proposal is validated on 45 benchmark time series databases and compared with several state-of-the-art methods, and obtains superior results in both earliness and accuracy. In addition, we show the practical applicability of our method for a real-world problem: the detection and identification of bird calls in a biodiversity survey scenario.

[1]  Jian Pei,et al.  A brief survey on sequence classification , 2010, SKDD.

[2]  Mark A. Girolami,et al.  vbmp: Variational Bayesian Multinomial Probit Regression for multi-class classification in R , 2008, Bioinform..

[3]  Paul Lukowicz,et al.  On general purpose time series similarity measures and their use as kernel functions in support vector machines , 2014, Inf. Sci..

[4]  Klaus Obermayer,et al.  Classification on Pairwise Proximity Data , 1998, NIPS.

[5]  Eamonn J. Keogh,et al.  Scalable Clustering of Time Series with U-Shapelets , 2015, SDM.

[6]  José Antonio Lozano,et al.  A general framework for the statistical analysis of the sources of variance for classification error estimators , 2013, Pattern Recognit..

[7]  Mohamed Medhat Gaber,et al.  A Survey of Classification Methods in Data Streams , 2007, Data Streams - Models and Algorithms.

[8]  J. Putter.  The Treatment of Ties in Some Nonparametric Tests , 1955 .

[9]  Juan José Rodríguez Diez,et al.  Early Fault Classification in Dynamic Systems Using Case-Based Reasoning , 2005, CAEPIA.

[10]  Roque Marín,et al.  Proceedings of the 11th Spanish association conference on Current Topics in Artificial Intelligence , 2005 .

[11]  Mohamed F. Ghalwash,et al.  Utilizing temporal patterns for estimating uncertainty in interpretable early decision making , 2014, KDD.

[12]  Philip S. Yu,et al.  Extracting Interpretable Features for Early Classification on Time Series , 2011, SDM.

[13]  Camelia Chira,et al.  Classifiers with a reject option for early time-series classification , 2013, 2013 IEEE Symposium on Computational Intelligence and Ensemble Learning (CIEL).

[14]  Miguel A. Labrador,et al.  A Survey on Human Activity Recognition using Wearable Sensors , 2013, IEEE Communications Surveys & Tutorials.

[15]  Latifur Khan,et al.  Feature Selection for Classification of Variable Length Multiattribute Motions , 2007 .

[16]  Mark A. Girolami,et al.  Bat Call Identification with Gaussian Process Multinomial Probit Regression and a Dynamic Time Warping Kernel , 2014, AISTATS.

[17]  Hyrum S. Anderson,et al.  Classifying with confidence from incomplete information , 2013, J. Mach. Learn. Res..

[18]  R. Scott Evans,et al.  Automated detection of physiologic deterioration in hospitalized patients , 2015, J. Am. Medical Informatics Assoc..

[19]  Eamonn J. Keogh,et al.  Time series shapelets: a new primitive for data mining , 2009, KDD.

[20]  Mark Girolami,et al.  Variational Bayesian Multinomial Probit Regression with Gaussian Process Priors , 2006, Neural Computation.

[21]  Eamonn J. Keogh,et al.  Experimental comparison of representation methods and distance measures for time series data , 2010, Data Mining and Knowledge Discovery.

[22]  J A Kogan,et al.  Automated recognition of bird song elements from continuous recordings using dynamic time warping and hidden Markov models: a comparative study. , 1998, The Journal of the Acoustical Society of America.

[23]  Hyrum S. Anderson,et al.  Reliable early classification of time series , 2012, 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[24]  Borja Calvo,et al.  scmamp: Statistical Comparison of Multiple Algorithms in Multiple Problems , 2016, R J..

[25]  Yong Duan,et al.  Early classification on multivariate time series , 2015, Neurocomputing.

[26]  J. Tobias,et al.  Threatened Birds of Asia: The BirdLife International Red Data Book. Collar, N.J., (Editor-in-chief), Andreev, A.V., Chan, S., Crosby, M.J., Subramanya, S. and Tobias, J.A. Maps by Rudyanto and M. J. Crosby. BirdLife International, Cambridge. 3,038 pages, in two volumes, £55.00. , 2001, Bird Conservation International.

[27]  Mohamed F. Ghalwash,et al.  Early classification of multivariate time series using a hybrid HMM/SVM model , 2012, 2012 IEEE International Conference on Bioinformatics and Biomedicine.

[28]  Philip S. Yu,et al.  Early classification on time series , 2012, Knowledge and Information Systems.

[29]  Janez Demsar,et al.  Statistical Comparisons of Classifiers over Multiple Data Sets , 2006, J. Mach. Learn. Res..

[30]  Claude Sammut,et al.  Classification of Multivariate Time Series and Structured Data Using Constructive Induction , 2005, Machine Learning.

[31]  Shuliang Wang,et al.  Data Mining and Knowledge Discovery , 2012, Springer Handbook of Geographic Information.

[32]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[33]  Latifur Khan,et al.  Real-time classification of variable length multi-attribute motions , 2006, Knowledge and Information Systems.

[34]  Rohit J. Kate.  Using dynamic time warping distances as features for improved time series classification , 2016, Data Mining and Knowledge Discovery.