Deep Gaussian Processes
Deep Gaussian Processes

Filter

Locations


Result types


Type of company


Industries


Company status

Number of employees

to

Founding year

to


Lock keywords

Exclude keywords

Optional keywords

Clear filters

Products and services for "Deep Gaussian Processes"

0 companies for "Deep Gaussian Processes"

Show More

„Deep Gaussian Processes“

Deep Gaussian Processes (DGPs) are probabilistic non-parametric models which extend the classic Gaussian Process (GP) by introducing multiple layers of latent variables. They allow for the combination of feature-engineering and learning of deep representations, providing an expressive and flexible model to represent complex data. A DGP consists of a sequence of GP layers, with the output of one layer becoming the input of the next. This makes it possible to learn complex non-linear correlations between data points. Additionally, the uncertainty of the predictions can be accurately estimated, making DGPs a valuable tool for Bayesian inference.