MATLAB File Help: dml.gridsearch

  gridsearch grid search method.
 
    DESCRIPTION
    This method can be used to optimize certain parameters of a
    multivariate analysis as specified in a crossvalidator object. The
    method will return an optimized multivariate method and hence can be
    used in a more complex multivariate analysis pipeline.
    Note: methods are supposed to handle optimization efficiently by warm starting
    at previous solutions. This also requires the 'restart' parameter of
    this method to be set to false. This is done to prevent unwanted use of
    previously estimated parameters by an object whenever 'restart' is true.
    
 
    EXAMPLE:
    X = rand(10,20); Y = [1 1 1 1 1 2 2 2 2 2]';
    v = dml.enet.lambdapath(X,Y,'logistic',5,1e-2);
    m = dml.gridsearch('validator',dml.crossvalidator('type','split','stat','accuracy','mva',dml.enet('type','logistic','restart',false)),'vars','L1','vals',v,'verbose',true);
    m = m.train(X,Y);
    Z = m.test(X);
 
    DEVELOPER
    Marcel van Gerven (m.vangerven@donders.ru.nl)

Superclasses	dml.method
Sealed	false
Construct on load	false

configs	the configurations (cartesian product of vals)
idx	the index of the method in the mva that is to be optimized
indims	dimensions of the input data (excluding the trial dim and time dim in time series data)
models	all models learned for each of the parameter settings
mva	optimized mva method
optimum	optimal configuration
outcome	outcome per configuration
restart	when false, starts at the previously learned parameters; needed for online learning and grid search
retrain	retrain the model using the optimal setting on all data
validator	the used cross-validator
vals	the values used
vars	the variables to optimize
verbose	whether or not to generate diagnostic output

	model	call the enclosed method
	test
	train