User-friendly surrogate model API

[sgbaird]

For example:

pip install matsci-opt-benchmarks

from matsci_opt import particle_packing

packing_fraction = particle_packing(
    mu1=1.2,
    mu2=1.5,
    mu3=2.3,
    std1=0.2,
    std2=0.1,
    std3=0.3,
    comp1=0.4,
    comp2=0.2,
    comp3=0.4,
    num_particles=500,
    safety_factor=1.5,
)

This could also be wrapped in a class to expose metadata associated with the function, such as the bounds of the training data that was used to produce the surrogate model.