we use Markove switching models when the parameters of the data generating process (DGP), over set of different unobserved states.Also when we don't know the current state of the (DGP ), but we can estimated the probability of each possible state.It used in economics in :
Asymmetrical behavior over (DGP)expension and recessions( Engle and Hamilliton,2010) ,Interest rate (Garica and Perron,1996), stock returns (Kime et al.1998)also in opidemiology ;in cidence rates of infictions diseases and non epidimic states (Lu et al 2010),and in psychology manic depressive state (Hamakev et al (2010),Enhance,we use with time series ,and any of parametrs(beta ,sigma,ARcomponents.