activity
20032022
most citedMulti-fidelity Bayesian Optimisation with Continuous Approximations

95 citations · 430 across the 31 of their papers we have counts for

collaborators
Showing 2017Show all

12 papers · 1 filter

astro-ph.CO201717 cited

Estimating Cosmological Parameters from the Dark Matter Distribution

Siamak Ravanbakhsh, Junier Oliva, Sebastien Fromenteau +4

A grand challenge of the 21st century cosmology is to accurately estimate the cosmological parameters of our Universe. A major approach to estimating the cosmological parameters is…

cs.LG201724 cited

A Generic Approach for Escaping Saddle points

Sashank J Reddi, Manzil Zaheer, Suvrit Sra +4

A central challenge to using first-order methods for optimizing nonconvex problems is the presence of saddle points. First-order methods often get stuck at saddle points, greatly d…

cs.LG20171 cited

Recurrent Estimation of Distributions

Junier B. Oliva, Kumar Avinava Dubey, Barnabas Poczos +2

This paper presents the recurrent estimation of distributions (RED) for modeling real-valued data in a semiparametric fashion. RED models make two novel uses of recurrent neural ne…

stat.ML201718 cited

Asynchronous Parallel Bayesian Optimisation via Thompson Sampling

Kirthevasan Kandasamy, Akshay Krishnamurthy, Jeff Schneider +1

We design and analyse variations of the classical Thompson sampling (TS) procedure for Bayesian optimisation (BO) in settings where function evaluations are expensive, but can be p…

cs.LG20174 cited

Data-driven Random Fourier Features using Stein Effect

Wei-Cheng Chang, Chun-Liang Li, Yiming Yang +1

Large-scale kernel approximation is an important problem in machine learning research. Approaches using random Fourier features have become increasingly popular [Rahimi and Recht,…

math.OC2017

Gradient Descent Can Take Exponential Time to Escape Saddle Points

Simon S. Du, Chi Jin, Jason D. Lee +3

Although gradient descent (GD) almost always escapes saddle points asymptotically [Lee et al., 2016], this paper shows that even with fairly natural random initialization schemes a…