Timezone: »
This paper introduces Wasserstein variational inference, a new form of approximate Bayesian inference based on optimal transport theory. Wasserstein variational inference uses a new family of divergences that includes both f-divergences and the Wasserstein distance as special cases. The gradients of the Wasserstein variational loss are obtained by backpropagating through the Sinkhorn iterations. This technique results in a very stable likelihood-free training method that can be used with implicit distributions and probabilistic programs. Using the Wasserstein variational inference framework, we introduce several new forms of autoencoders and test their robustness and performance against existing variational autoencoding techniques.
Author Information
Luca Ambrogioni (Donders Institute)
Umut Güçlü (Donders Institute for Brain, Cognition and Behaviour, Radboud University)
Yağmur Güçlütürk (Donders Institute for Brain, Cognition and Behaviour, Radboud University)
Max Hinne (University of Amsterdam)
Marcel A. J. van Gerven (Radboud Universiteit)
Eric Maris (Donders Institute)
More from the Same Authors
-
2020 Poster: GAIT-prop: A biologically plausible learning rule derived from backpropagation of error »
Nasir Ahmad · Marcel A. J. van Gerven · Luca Ambrogioni -
2020 Spotlight: GAIT-prop: A biologically plausible learning rule derived from backpropagation of error »
Nasir Ahmad · Marcel A. J. van Gerven · Luca Ambrogioni -
2017 Poster: GP CaKe: Effective brain connectivity with causal kernels »
Luca Ambrogioni · Max Hinne · Marcel Van Gerven · Eric Maris -
2017 Poster: Reconstructing perceived faces from brain activations with deep adversarial neural decoding »
Yağmur Güçlütürk · Umut Güçlü · Katja Seeliger · Sander Bosch · Rob van Lier · Marcel A. J. van Gerven -
2016 Demonstration: Biometric applications of CNNs: get a job at "Impending Technologies"! »
Sergio Escalera · Isabelle Guyon · Baiyu Chen · Marc Quintana · Umut Güçlü · Yağmur Güçlütürk · Xavier Baró · Rob van Lier · Carlos Andujar · Marcel A. J. van Gerven · Bernhard E Boser · Luke Wang -
2016 Poster: Brains on Beats »
Umut Güçlü · Jordy Thielen · Michael Hanke · Marcel van Gerven · Marcel A. J. van Gerven