A method of finding frequently occurring patterns in spike trains is proposed. Due to stochastic fluctuation, patterns do not appear in exactly the same way. Therefore spike train subsequences are clustered using a similarity measure based on a positive definite kernel for alignments of local interval structures. By applying the method to synthetic data, spike trains generated from different neuron types are classified correctly. For synthetic and real data, patterns that repeatedly appear were successfully extracted.
CITATION STYLE
Tezuka, T. (2015). Spike train pattern discovery using interval structure alignment. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9490, pp. 241–249). Springer Verlag. https://doi.org/10.1007/978-3-319-26535-3_28
Mendeley helps you to discover research relevant for your work.