摘要: Continuous-time autoregressive (CAR) processes have been of interest to physicists and engineers for many years (see e.g., Fowler, 1936 ). Early papers dealing with the properties statistical analysis such processes, more general continuous-time moving average (CARMA) include those Doob (1944) , Bartlett (1946) Phillips (1959) Durbin (1961) . In last ten there has a resurgence in partly as result very successful application stochastic differential equation models problems finance, exemplified by derivation Black-Scholes option-pricing formula its generalizations ( Hull White, 1987 Numerous examples econometric applications are contained book Bergstrom (1990) also utilized successfully modelling irregularly-spaced data Jones, 1981 1985 Jones Ackerson At same time an increasing realization that non-linear series provide much better representations empirically observed than linear models. The threshold ARMA Tong, 1983 1990 particularly representing wide variety sets, ARCH GARCH Engle (1982) Bollerslev (1986) respectively had great success financial data. versions developed Nelson this paper we discuss models, their basic properties, relationship discrete-time inference based on observations made at discrete times which analogues Tong's