System Identification with Analog and Counting Process Observations I: Hybrid Stochastic Intensity and Likelihood.

In a number of application areas there is growing interest in system identification for systems whose observation processes consist of both analog and counting process signals. But so far few system identification techniques exist for these cases and likelihood functions have so far not been available. Here we introduce a new hybrid stochastic intensity and use it to construct, for the first time, an analog-counting process likelihood.