We propose the Bayesian information criterion (BIC) and the Akaike information criterion (AIC) for model selection in hidden Markov models (HMM) when the number of states is unknown. The exact ...

Probability and Statistics Group research at the School of Mathematical and Computer Sciences at Heriot-Watt University, Edinburgh.

For additive actuator and sensor faults, we propose a systematic method to design a state-space fault estimation filter directly from Markov parameters identified from fault-free data. We address this ...