This PR adds a new surrogate namely the LinearDeterministicSurrogate which is just a linear model of type f(x) = mx +b, where are m are the coefficients and b is the intercept. Tis is useful for multi-objective optimization with partially known objective functions (for example the number of parameters of a Neural Network in the context of hyperparameter optimization).
This PR adds a new surrogate namely the
LinearDeterministicSurrogate
which is just a linear model of typef(x) = mx +b
, where arem
are the coefficients andb
is the intercept. Tis is useful for multi-objective optimization with partially known objective functions (for example the number of parameters of a Neural Network in the context of hyperparameter optimization).