AFL.agent.HscedGaussianProcess module#
- class AFL.agent.HscedGaussianProcess.DummyGP(dataset, kernel=None)#
Bases:
object- predict(components)#
- class AFL.agent.HscedGaussianProcess.GPR(dataset, inputs=None, uncertainties=None, kernel=None, heteroscedastic=False)#
Bases:
object- construct_data(heteroscedastic=True, preprocess=True)#
- optimize(N=1000, gamma=0.5, final_monitor_step=None, progress_bar=False, tol=0.0001)#
- predict(components)#
- reset_GP(dataset, kernel=None, heteroscedastic=False)#
- reset_monitoring(log_dir='test/', iter_period=1)#
- transform_domain(components=None)#
- class AFL.agent.HscedGaussianProcess.HeteroscedasticGaussian(**kwargs)#
Bases:
LikelihoodDefines the heteroscedastic gaussian likelihood