Learning Comprehensible Descriptions of Multivariate Time Series

Supervised classiication is one of the most active areas of machine learning research. Most work has focused on classiication in static domains, where an instantaneous snapshot of attributes is meaningful. In many domains, attributes are not static; in fact, it is the way they vary temporally that can make classiication possible. Examples of such domains include speech recognition, gesture recognition and electrocardiograph clas-siication. While it is possible to use ad hoc, domain-speciic techniques for \\atten-ing" the time series to a learner-friendly representation , this fails to take into account both the special problems and special heuris-tics applicable to temporal data and often results in unreadable concept descriptions. Though traditional time series techniques can sometimes produce accurate classiiers, few can provide comprehensible descriptions. We propose a general architecture for classiica-tion and description of multivariate time series. It employs event primitives to analyse the training data and extract events. These events are clustered, creating proto-typical events which are used as the basis for creating more accurate and comprehensi-ble classiiers. A minimal implementation of this architecture, called TClass, is applied to two domains, one real and one artiicial and compared against a na ve approach. TClass shows great promise, particularly in compre-hensibility, but also in accuracy.