Pattern Extraction for Time Series Classi

In this paper, we propose some new tools to allow machine learning classiiers to cope with time series data. We rst argue that many time-series classiication problems can be solved by detecting and combining local properties or patterns in time series. Then, a technique is proposed to nd patterns which are useful for classiication. These patterns are combined to build interpretable classiication rules. Experiments , carried out on several artiicial and real problems, highlight the interest of the approach both in terms of interpretability and accuracy of the induced classiiers.