Skip to main content

Table 9 Overview of the parameter space screened to tune the prior of the parameters from a Bayesian Logistic regression, which was used as the last layer in the HBLL models. More precisely, the precision of the Gaussian prior of the model parameters was optimized using the validation set

From: Achieving well-informed decision-making in drug discovery: a comprehensive calibration study using neural network-based structure-activity models

Model

Hyperparameter

Explored Space

HBLL

Precision of the

{100, 150, 200, 250, 300, 400, 600,

Gaussian prior

800, 1000, 1200, 1400, 1600, 1800, 2000}

Back to article page

Journal of Cheminformatics

ISSN: 1758-2946

Contact us