没看完¶

bayesian nonparametric models such as Gaussian processes

function extrapolation problems

kernel learning framework

** reverse the human-like and inductive biases of human across a set of behavioral experiments

to gain psychological insights and to extrapolate in human

model ability determined by

1. its support(which solutions are a priori possible )
2. inductive biases (which solutions are a priori likely)

controlled by a covariance kernel