Set Parameters (AI Studio Core)

Description

The Set Parameters operator takes a set of parameters as input. Operators like Optimize Parameters (Grid) or Read Parameters can be used as a source of parameter set. The Set Parameters operator takes this parameter set and assigns these parameter values to the parameters of the specified operator which can be specified through the name map parameter. This menu has two columns: 'set operator name' and 'operator name'. The 'set operator name' column is used for specifying the name of the operator (or the name of the parameter set) that generated this parameter set and the 'operator name' column specifies the name of the target operator i.e. the operator that will use this parameter set.

This operator can be very useful, e.g. in the following scenario. If one wants to find the best parameters for a certain learning scheme, one is usually also interested in the model generated with these optimal parameters. Finding the best parameter set is easily possible by operators like the Optimize Parameters (Grid) operator. But generating a model with these optimal parameters is not possible because such parameter optimization operators do not return the IOObjects produced within, but only a parameter set and a performance vector. This is, because the parameter optimization operators know nothing about models, but only about the performance vectors produced within and producing performance vectors does not necessarily require a model. To solve this problem, one can use the Set Parameters operator. Usually, a process with the Set Parameters operator contains at least two operators of the same type, typically a learner. One learner may be an inner operator of a parameter optimization scheme and may be named 'Learner', whereas a second learner of the same type named 'OptimalLearner' comes after the parameter optimization scheme and should use the optimal parameter set returned by the optimization scheme. The Set Parameters operator takes the parameter set returned by the optimization scheme as input. The name 'Learner', is specified in the 'set operator name' column and the name 'OptimalLearner' is specified in the 'operator name' column of the name map parameter. Each parameter in this list maps the name of an operator that was used during the optimization (in our case this is 'Learner') to an operator that should now use these parameters (in our case this is 'OptimalLearner'). An important thing to note here is the sequence of operators. The 'Learner', operator should be first in sequence, followed by the Set Parameters operator and finally the 'OptimalLearner' operator.