Model Identification for a Parameter

m = Model

d = DataSimulator.simulate(m0)

for ( m : ModelGenerator.generate (m0) ) {

      l = LHood ( d with model m )

      list.add( l );

}

Valid ID

1. Model has ID iff, l is different for every m
2. MLE is viable if mle of repeated data d includes m0 within 6-sigma

Output

1. m0
2. Plot of MLE. Ave. mle &  standard error. The particular plot to choose depends on the model parameter under investigation (and also the model simulator) can be decided at the job level.

Input

1. Initial model m0
2. Data simulator algorithm (dynamic)
3. Model generator algorithm (specific to parameter) (dynamic)
4. LHood algorithm (dynamic)