Max Welling

Max Welling

A Dutch machine learning researcher at the University of Amsterdam, co-inventor with Diederik Kingma of the Variational Autoencoder, which framed generative modelling as learning a compressed probabilistic latent space from which new samples can be drawn. With Thomas Kipf he also introduced graph convolutional networks, extending deep learning to data structured as networks rather than grids or sequences. Much of his recent work applies machine learning to physics and chemistry simulation. (See also: Diederik Kingma, Generative AI, Neural network, Embeddings)