Long Tran-Thanh

Professor. Does exist at Warwick Uni. In theory: working on AI/ML - bandits, game theory, and AI4SG. In practice: procrastinates & drinks craft beers.

[Long Tran-Thanh @ltt_hal]

A solely technology-focused approach would definitely fail. I have been speaking with many policy makers, business leaders, and technologies in the last few months, and it's quite worrisome that not many of them really understood this at all. It's time to change this mindset! 2/2

[Long Tran-Thanh @ltt_hal]

Why AI typically fails in real life (so far)? Many people don't understand that applied AI requires interdisciplinary approach. It only works in real life if it's problem-driven and the solution is co-created from the very beginning by people from multiple disciplines 1/n

[Long Tran-Thanh @ltt_hal]

Wow, Turing Award 2024 for Reinforcement Learning. A long awaited and deserved recognition:

https://t.co/JlRMNlu1oo

[Long Tran-Thanh @ltt_hal]

Many congrats Martin and Sayan!

[Department of Computer Science - Warwick @WarwickDCS]

A paper co-authored by Matthias C. Caro has been featured on the cover of PRX Quantum, a premier journal for quantum information science and technology research ⚛️

More info: https://t.co/qkizNwx73u

Image credit: APS/YOHOHO

[Long Tran-Thanh @ltt_hal]

Welcome Arnab to Warwick! Looking fwd to work with you on many things :)

[Arnab Bhattacharyya @abhatt2]

Personal Update: I've moved to the University of Warwick, UK 🇬🇧.

It was a tremendous 5 years in Singapore @NUSComputing, and my move was bittersweet.

But I'm so very excited to be part of Warwick (and UK academia more broadly). Stay tuned for more updates!

[Department of Computer Science - Warwick @WarwickDCS]

We are happy to announce that Dr Sayan Bhattacharya has been awarded a €2million ERC Consolidator grant for a 5-year project entitled "Towards a Dynamic Algorithms Centric Theory of Linear Programming" 👏

More info: https://t.co/s9PJAa8mma

[Long Tran-Thanh @ltt_hal]

So there’s hope that RL is back in the game.

[Richard Sutton @RichardSSutton]

The original RL algorithms, inspired by natural learning, were online and incremental—they were streaming in the sense that they learned from each increment of experience as it happened, then discarded it, never to be processed again. The streaming algorithms were simple and elegant, but the first big successes of RL in deep learning were not with streaming algorithms. Instead, methods such as DQN chopped the stream of experience into individual transitions, then stored and sampled them in arbitrary batches. Subsequent work followed, extended, and refined the batch approach into asynchronous and offline RL, while the streaming approach languished, unable to produce good results in popular deep learning domains.

Until now. Now researchers at the University of Alberta have shown that streaming RL algorithms can work just as well as DQN on Atari and Mujoco tasks (https://t.co/S4D6lSvdxz). How did they do it? Mostly just by getting signal normalization and step-size bounding right for the streaming case—otherwise they use standard streaming algorithms like TD(lambda) and Q(lambda). To me it looks like they were simply the first researchers knowledgeable of streaming RL algorithms to seriously address deep RL without being over-influenced by batch-oriented software and batch-oriented supervised-learning ways of thinking.

[Richard Sutton @RichardSSutton]

The original RL algorithms, inspired by natural learning, were online and incremental—they were streaming in the sense that they learned from each increment of experience as it happened, then discarded it, never to be processed again. The streaming algorithms were simple and elegant, but the first big successes of RL in deep learning were not with streaming algorithms. Instead, methods such as DQN chopped the stream of experience into individual transitions, then stored and sampled them in arbitrary batches. Subsequent work followed, extended, and refined the batch approach into asynchronous and offline RL, while the streaming approach languished, unable to produce good results in popular deep learning domains.

Until now. Now researchers at the University of Alberta have shown that streaming RL algorithms can work just as well as DQN on Atari and Mujoco tasks (https://t.co/S4D6lSvdxz). How did they do it? Mostly just by getting signal normalization and step-size bounding right for the streaming case—otherwise they use standard streaming algorithms like TD(lambda) and Q(lambda). To me it looks like they were simply the first researchers knowledgeable of streaming RL algorithms to seriously address deep RL without being over-influenced by batch-oriented software and batch-oriented supervised-learning ways of thinking.

[Long Tran-Thanh @ltt_hal]

Today we have Ana Stoica from @MPI_IS speaking at our FAIS seminar on socially-aware ML and optimization.

[FoundationsAISeminars @FAIS_Warwick]

Today, we’re excited to host Ana-Andreea Stoica from @MPI_IS at our Foundations of AI Seminar Series,@FAIS_Warwick!

🗓️ Nov 11th
🕐Time: 2pm (GMT)
📍 Zeeman Building, MS.03

@warwickdcs
@uniofwarwick
#WarwickAI

[FoundationsAISeminars @FAIS_Warwick]

Today, we’re excited to host Ana-Andreea Stoica from @MPI_IS at our Foundations of AI Seminar Series,@FAIS_Warwick!

🗓️ Nov 11th
🕐Time: 2pm (GMT)
📍 Zeeman Building, MS.03

@warwickdcs
@uniofwarwick
#WarwickAI

[Long Tran-Thanh @ltt_hal]

Oh wow, first we have lottery ticket hypothesis, now we have lottery ticket registrations. What else will the ML community bring to us in the future?

[NeurIPS Conference @NeurIPSConf]

Due to a high demand for registrations, NeurIPS will be moving towards a randomized lottery system, effective immediately. Authors of accepted conference and workshop papers are still guaranteed registration, but this may change as we release spots to the lottery, so we urge authors to register ASAP.

Read more:
https://t.co/BhCBy7rfhz

[Long Tran-Thanh @ltt_hal]

@yalidux @arxiv Indeed, the superiority of procrastinating:)

[Long Tran-Thanh @ltt_hal]

As @arxiv is down, first time ever appreciate the fact that I have 40+ Chrome tab open with different arxiv pdfs and don’t have to reload them. Now it’s time to read them (haven’t read at least 75% of them yet tbh)

[Long Tran-Thanh @ltt_hal]

The flagship conference in knowledge representation and reasoning #KR2024 will be in Hanoi, Vietnam between 2-8Nov 2024. Keynote speakers are Meghyb Bienvenu, @rao2z, Nina Narodytska, @SheilaMcIlraith, and Murray Shanahan. Join us to enjoy both the conf and the cultural heritage.

[Long Tran-Thanh @ltt_hal]

Warwick DCS has 8 papers accepted to @NeurIPSConf this year. In the last 2 years we have been heavily investing in #FoundationOfAI, building on top of our renowned reputation in theoretical research in CS and AI, and will be hiring more! Join us if you’re into theoretical AI/ML

[Department of Computer Science - Warwick @WarwickDCS]

Researchers from the Department of Computer Science at the University of Warwick have had eight papers accepted for publication at NeurIPS 2024! Congratulations to all involved 👏 Read more: https://t.co/gXG0VLMF5e

[Long Tran-Thanh @ltt_hal]

@CsabaSzepesvari Yeah, it’s a very unfortunate, myopic and populist reaction. The good news is that I just got my ETA for 5 years. Now I can come and visit you at some point :)

[Long Tran-Thanh @ltt_hal]

This is insane! My PhD student, who got 2 accepted first-authored papers to #NeurIPS2024, cannot attend the conf because the visa processing time in Canada is 10 months (sic!). What’s happening in Canada? It’s time to look for better accessible places to host Neurips!!

[Department of Computer Science - Warwick @WarwickDCS]

We are seeking PhD candidates in the topic of Multiagent Systems and related areas, with particular emphasis on one or more of: computational social choice, algorithmic game theory, multiagent learning, and social and economic networks. More info: https://t.co/VymGKyYmY7

[Long Tran-Thanh @ltt_hal]

@mehtalearner Maybe with a Taylor Swift ticket one could get it faster ;) I heard that she’s playing at the same time in Vancouver as neurips

[Long Tran-Thanh @ltt_hal]

@JohnLPoquiz True! For icml this year which was in Vienna, my student got his visa interview date to be after the conference (sic!)