Ruslan Salakhutdinov

Ruslan Salakhutdinov is a professor in the Machine Learning Department at Carnegie Mellon University. He earned a doctorate at the University of Toronto in 2009, working with Geoffrey Hinton. His research investigates statistical machine learning, deep generative models, and methods for learning useful representations from data.

With Hinton, Salakhutdinov co-authored the 2006 paper Reducing the Dimensionality of Data with Neural Networks. It showed how multilayer networks could learn compact codes that reconstruct high-dimensional inputs. The work contributed to the renewed interest in deep neural networks and the methods used to train them.

Salakhutdinov’s wider work has included probabilistic models and learning approaches that investigate how systems can generalize from limited examples. He also served as Apple’s director of AI research before returning to academic work. His contribution connects the development of model capabilities with statistical explanations of representation, uncertainty, and inference, alongside teaching and supervision of later researchers.