Open Access

Table A.2

Overview of the GP model’s components.

Property Value
GP model type Exact, Hadamard Multitask
Feature extractor CNN
Feature normalization Scale to bounds (-1, 1)
Likelihood function Gaussian likelihood
Mean function Constant
Covariance function/kernel Matern kernel (smoothness
0.5)
Task covariance function/kernel Index kernel (rank 3)
Optimizer Adam
Learning rate 0.1
Loss function Marginal log-likelihood
Training iterations 150

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.