Innovation method
In statistics [1] , the Innovation Method provides an estimator[2] for the parameters[3] of stochastic differential equations[4] given a time series[5] of (potentially noisy[6] ) observations[7] of the state variables[8] . In the framework of continuous-discrete state space models, the innovation estimator is obtained by maximizing the log-likelihood[9] of the corresponding discrete-time[10] innovation process with respect to the parameters. The innovation estimator can be classified as a M-estimator[11], a quasi-maximum likelihood estimator [12] or a prediction error[13] estimator depending of the inferential considerations that want to be emphasized.
- ^ https://en.wikipedia.org/wiki/Statistics
- ^ https://en.wikipedia.org/wiki/Estimator
- ^ https://en.wikipedia.org/wiki/Parameters
- ^ https://en.wikipedia.org/wiki/Stochastics_differential_equation
- ^ https://en.wikipedia.org/wiki/Time_series
- ^ https://en.wikipedia.org/wiki/Ramdomness
- ^ https://en.wikipedia.org/wiki/Observation
- ^ https://en.wikipedia.org/wiki/State_variable
- ^ https://en.wikipedia.org/wiki/Maximun_likelihood_estimation
- ^ https://en.wikipedia.org/wiki/Discrete_time_and_continuous_time
- ^ https://en.wikipedia.org/wiki/M-estimator
- ^ https://en.wikipedia.org/wiki/Quasi-maximum_likelihood_estimate
- ^ https://en.wikipedia.org/wiki/Mean_squared_prediction_error