Hung Tran-The

Lecturer at Applied Artificial Intelligence Insititute, Deakin University. I work on Bandit, Reinforcement Learning, Bayesian Optimization.

[Gabriel Peyré @gabrielpeyre]

Bregman divergences are convex distance-like functionals which are locally Euclidean. Most algorithm handling Euclidean distances generalize to Bregman divergences. https://t.co/WSZqCaaIl2

[Yufei Zhao @yufeizhao]

Congratulations to Jinyoung Park and Huy Pham for proving the Kahn-Kalai conjecture---a central open problem in probabilistic combinatorics. Truly exciting breakthrough! https://t.co/BZxaceCDW7
The story of Jinyoung's extraordinary path to mathematics:
https://t.co/C7ADGgiKTV

[eytan bakshy @eytan]

Bayesian optimization is often described as a way of taking the human out of the loop for time consuming multi-dimensional optimization problems. But in many practical applications like policy optimization, decision makers have difficulty formulating an objective. 1/n #AISTATS

[Houssam Zenati @HoussamZenati]

🎉Happy to announce our work https://t.co/f8xCKphOww on Efficient Kernel UCB for contextual bandits! 🎉 Join us at the poster session on Wed at 3:30 UTC #AISTATS2022 We improve the computational efficiency of the algorithm while recovering the same regret as the standard methods

[Siu Lun Chau @Chau9991]

Happy to announce our paper "Learning Inconsistent Preferences with Gaussian Processes" has been accepted to #AISTATS2022.

Interested in modelling inconsistent preferences? Come to our poster session 2 today at 6:15 p.m. BST!

https://t.co/Mim1Y9thmH

@javiergonzh @sejDino

[Jonathan Scarlett @j_m_scarlett]

Myself and others are organising a special issue for IEEE Journal on Selected Areas in Information Theory, on "Deep Learning Methods for Inverse Problems" - submission date is May 15.
https://t.co/sBDzoUoVAD

[Louis Tiao @louistiao]

A Gaussian process is a collection of (infinitely many) random variables such that the marginal distribution over any finite subset is a multivariate Gaussian distribution.

This somewhat arcane description is easier to conceptualize when we focus on a subset of two variables:

[Gabriel Peyré @gabrielpeyre]

Natural neighbor interpolation (Robin Sibson) is a generalization of piecewise linear interpolation to higher dimensions. https://t.co/ZoYLPC6FIA https://t.co/R0aaz8xpR5

[Ilija Bogunovic @ilijabogunovic]

Looking to hire a Ph.D. student to work on algorithmic data-driven decision making (Topics: Multi-agent learning, RL, Bayesian optimization, Experiment design) at UCL E&EE @ucleeenews. Retweets appreciated!

Details: https://t.co/gkVdsB5ymJ

[Gabriel Peyré @gabrielpeyre]

The push-forward operator is a linear map between measures which operates by displacement of the support of the points. Corresponds to the composition of random vector by a function, and to the change of variable formula of integration. https://t.co/qwksw0EOEz

[Gabriel Peyré @gabrielpeyre]

Monte Carlo integration approximates integrals at a rate 1/sqrt(n) independent of the dimension. https://t.co/hqEK1IHrmy

[Aurelien Lucchi @AurelienLucchi]

While some of us are doing extra work over the weekend for ICLR and/or Aistats...

[AISTATS Conference @aistats_conf]

AISTATS 2022 is still accepting submissions! The deadline for abstracts is 1 week away - Friday, 8 October 2021 11:59am (UTC). And the deadline for papers is 2 weeks away - Friday, 15 October 2021 11:59am (UTC). 1/2

[Preetum Nakkiran @PreetumNakkiran]

Reviewer 2 complains it has "too many ideas".
@prfsanjeevarora says "gave me some new ideas".
This Friday at Noon PT… come hear about the paper that reviewers don't want you to read! https://t.co/bjA3cXqfpU

I'm speaking at @ml_collective's DLCT: https://t.co/UUDjstsSWz

[Silvia Chiappa @csilviavr]

A position is still available. For informal queries, please reach out to me by email.
Retweets appreciated!

[Gabriel Peyré @gabrielpeyre]

Fourier descriptors (Zahn and Roskies, 1972) are normalized Fourier moments features which are invariant to translation and rotation. https://t.co/e161gLtlYA

[Gabriel Peyré @gabrielpeyre]

Fokker–Planck equation equivalently describes the movement of a random particule with a drift (as a stochastic ODE) and the evolution of its density (as a PDE). https://t.co/F8up4bYD2d

[Konstantin Mishchenko @konstmish]

Relative smoothness is one of the most reinvented definitions in optimization. It gives a simple condition for convergence of mirror descent and dates back at least to 2010. Its many applications should be the way mirror descent is motivated in lectures.
https://t.co/N0hFjzZA0O

[Dr. Ganapathi Pulipaka 🇺🇸 @gp_pulipaka]

... The Human Brain and the Universe Have Same Design and Neural Network Structures with 100 Billion Neurons & 100 Billion Galaxies. #DataScience #MachineLearning #IoT #IIoT #Python #RStats #JavaScript #ReactJS #Serverless #Programming #100DaysofCode
https://t.co/rYLYHwgOIv

[MathType @MathType]

Today is the 202nd birthday anniversary of Sir George Gabriel Stokes, a big name in both #Physics and #Math. One of the theorems that bears his name in #VectorCalculus is a theorem so powerful that it contains many special cases that are themselves powerful theorems. #MathType