Handling Selective Participation in Neuron Assembly Detection

With the objective to detect neuron assemblies in recorded parallel spike trains, we develop methods to find frequent parallel episodes in parallel point processes (or event sequences) that allow for imprecise synchrony of the events constituting occurrences (temporal imprecision) as well as incomplete occurrences (selective participation). The temporal imprecision problem is tackled by frequent pattern mining using two different notions of synchrony: a binary notion that captures only the number of instances of a pattern and a graded notion that captures both the number of instances as well as the precision of synchrony of its events. To cope with selective participation, which is the main focus of this paper, a reduction sequence of items (or event types) is formed based on found frequent patterns and guided by pattern overlap, for which we explore different concept. We demonstrate the performance of our methods on a large number of (artificially generated) data sets with injected parallel episodes, which mimic actually recorded parallel spike trains.