Thesis proposal : A study on the discriminatory capacity of the temporal information on supervised time series classification problems

Time series classification has been always categorized as a particular case of classification in which the input objects are ordered sequences. As such, the research community has assumed that specific methods are required for dealing with this type of data, without really analysing this hypothesis. Specific methods are usually computationally expensive or demand some semantic treatment of the series that may turn the method cumbersome. In this thesis time series classification is addressed from a new point of view: the discriminatory power of the temporal information. In other words, given a dataset, we want to analyse the relevance of the temporal information (the order of the elements of a series, for instance) for classification and explore in which cases the specific methods are necessary and why in depth. Departing from distance based time series classification, the goal of this thesis is to explore which are the temporal characteristic of the time series data in order to measure how relevant they are for classification.

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

[2]  Clu-istos Foutsos,et al.  Fast subsequence matching in time-series databases , 1994, SIGMOD '94.

[3]  Marcella Corduas,et al.  Time series clustering and classification by the autoregressive metric , 2008, Comput. Stat. Data Anal..

[4]  José Antonio Lozano,et al.  A review on distance based time series classification , 2018, Data Mining and Knowledge Discovery.

[5]  Donald J. Berndt,et al.  Using Dynamic Time Warping to Find Patterns in Time Series , 1994, KDD Workshop.

[6]  Gareth J. Janacek,et al.  A Run Length Transformation for Discriminating Between Auto Regressive Time Series , 2014, J. Classif..

[7]  Padhraic Smyth,et al.  Clustering Sequences with Hidden Markov Models , 1996, NIPS.

[8]  Carlos Agón,et al.  Time-series data mining , 2012, CSUR.

[9]  Li Wei,et al.  Fast time series classification using numerosity reduction , 2006, ICML.

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

[11]  Stephan K. Chalup,et al.  GDTW-P-SVMs: Variable-length time series analysis using support vector machines , 2013, Neurocomputing.

[12]  Renée J. Miller,et al.  Similarity search over time-series data using wavelets , 2002, Proceedings 18th International Conference on Data Engineering.

[13]  Sylvie Gibet,et al.  On Recursive Edit Distance Kernels With Application to Time Series Classification , 2010, IEEE Transactions on Neural Networks and Learning Systems.

[14]  Eamonn J. Keogh,et al.  On the Need for Time Series Data Mining Benchmarks: A Survey and Empirical Demonstration , 2002, Data Mining and Knowledge Discovery.

[15]  Daniel T. Larose,et al.  An Introduction to Data Mining , 2005 .