A Wrapped Normal Distribution on Hyperbolic Space for Gradient-Based Learning
Yoshihiro Nagano · Shoichiro Yamaguchi · Yasuhiro Fujita · Masanori Koyama

Wed Jun 12th 12:00 -- 12:05 PM @ Hall A

Hyperbolic space is a geometry that is known to be well-suited for representation learning of data with an underlying hierarchical structure. In this paper, we present a novel hyperbolic distribution called \textit{pseudo-hyperbolic Gaussian}, a Gaussian-like distribution on hyperbolic space whose density can be evaluated analytically and differentiated with respect to the parameters. Our distribution enables the gradient-based learning of the probabilistic models on hyperbolic space that could never have been considered before. Also, we can sample from this hyperbolic probability distribution without resorting to auxiliary means like rejection sampling. As applications of our distribution, we develop a hyperbolic-analog of variational autoencoder and a method of probabilistic word embedding on hyperbolic space. We demonstrate the efficacy of our distribution on various datasets including MNIST, Atari 2600 Breakout, and WordNet.

Author Information

Yoshihiro Nagano (The University of Tokyo)
Shoichiro Yamaguchi (Preferred Networks)
Yasuhiro Fujita (Preferred Networks, Inc.)
Masanori Koyama (Preferred Networks Inc. )

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors