Learning First Order Logic Time Series Classifiers

A method for learning multivariate time series classifiers by inductive logic programming is presented. Two types of background predicate that are suited for this task are introduced: interval based predicates, such as "always", and distance based, such as the euclidean distance. Special purpose techniques are presented that allow these predicates to be handled efficiently when performing top-down induction. Furthermore, by employing boosting, the accuracy of the resulting classifiers can be improved significantly. Experiments on several different datasets show that the proposed method is highly competitive with previous approaches. On all data sets, the proposed method achieves better than all previously reported results. Moreover, the comprehensibility of the induced classifiers is found to be better than for classifiers produced by previous approaches.

• Type: Conference
• Year: 2000
• Where: ILP
• Authors: Juan José Rodríguez, Carlos J. Alonso, Henrik Boström
• Keywords: Automated Reasoning, ILP 2000, Inductive Logic Programming, Interval Based Predicates, Time Series Classifiers