- Lisa_teatalk - mailman.iro.umontreal.ca

Re: [Lisa_teatalk] [Tea Talk] Paul Cisek (Udem ) Fri 19 July 2019 10h30 Mila Agora
by Rim Assouel 19 Jul '19

19 Jul '19

Reminder : this is in 45min ! > On Jul 15, 2019, at 9:14 AM, Rim Assouel <rim.assouel(a)gmail.com> wrote: > > This week we have Paul Cisek from Udem giving a talk on Rethinking behavior from an evolutionary perspective at 10h30 in room Mila Agora. > > Will this talk be streamed <https://mila.bluejeans.com/4862024040/webrtc>? Yes > > Meeting with Paul : if you would like to meet with the speaker in the afternoon after the talk please fill this google form in ASAP : https://forms.gle/85xGyQMy2o3mHHFi9 <https://forms.gle/85xGyQMy2o3mHHFi9> > > > ############### > ####### > ### > # > > Non-Mila members mailing list : Reminder that a mailing list for non-mila people who would like to get tea talks notifications has been set up : https://groups.google.com/forum/#!forum/mila-teatalks <https://groups.google.com/forum/#!forum/mila-teatalks> > > Internal website : Tea Talks recordings are available here ( you need to sign in with your Mila account) : https://sites.google.com/lisa.iro.umontreal.ca/tea-talk-recordings/home?aut… <https://sites.google.com/lisa.iro.umontreal.ca/tea-talk-recordings/home?aut…> > > # > ### > ####### > ############### > See you there! > The Tea Talk Team > > TITLE Rethinking behavior from an evolutionary perspective > > ABSTRACT > In both AI research and psychological theory, the human brain is usually thought of as an information processing system that encodes and manipulates representations of knowledge to produce plans of action. This view leads to a decomposition of behavior into putative functions such as object recognition, memory, decision-making, action planning, etc., inspiring the search for the neural correlates of these functions and attempts to simulate them in computational models. However, empirical neurophysiological data does not support many of the predictions of these classic subdivisions, consistently showing divergence and broad distribution of putatively unified functions, “mixed representations” that combine sensory, motor, and cognitive variables in single neurons, and a general incompatibility with the conceptual subdivisions posited by psychological theories. In this talk, I will explore the possibility of resynthesizing a different set of functional subdivisions, guided by the growing body of data on the evolutionary process that produced the human brain. The main part of my talk will summarize, in chronological order, the sequence of innovations that appeared in nervous systems along the lineage that leads from the earliest multicellular animals to humans. Along the way, functional subdivisions and elaborations will be introduced in parallel with the neural specializations that made them possible, gradually building up an alternative conceptual taxonomy of brain functions. These functions emphasize mechanisms for real-time interaction with the world, rather than for building explicit knowledge of the world, and the relevant representations emphasize pragmatic outcomes rather than decoding accuracy, mixing variables in just the way seen in real neural data. I will argue that this alternative taxonomy better delineates the real functional pieces into which the human brain is organized, offers a more natural mapping to real neural structures, and provides a better set of constraints for constructing artificial systems aimed at replicating brain function > > BIO > Paul Cisek was born in Poland and emigrated to the US as a child. While he was originally trained in computer science and artificial intelligence, he became more interested in natural intelligence and enrolled in graduate studies under the supervision of Stephen Grossberg and Daniel Bullock and Boston University, obtaining a PhD in computational neuroscience. He then became interested in empirical neuroscience, and joined the laboratories of Stephen Scott and John Kalaska to study the processes of selection, planning, and control of actions by recording neural activity in the cerebral cortex of monkeys. Since then, he has established his own laboratory at the University of Montreal, in which he combines all of these techniques to investigate the cortical and subcortical mechanisms controlling our interactions with the world.

1 0

Re: [Lisa_teatalk] [mila-tous] [Tea Talk] Dan Sodickson (NYU) Fri 17 May 2019 10h30 Auditorium H2
by Joseph Paul Cohen 14 Mai '19

14 Mai '19

Can we please use a public calendar? After all the debates about the "right way" to do calendars mine does not show that we have a tea talk on Friday. On Tue, May 14, 2019 at 11:47 AM Rim Assouel <rim.assouel(a)gmail.com> wrote: > > This week we have Dan Sodickson from NYU giving a talk on Machine Learning and Medicine: How AI will change the way we see patients, and the way we see ourselves. at 10h30 in Auditorium H2. > > Auditorium H2 is on the 2nd floor of the 6650 building. Indications will be put from the agora and the elevators as well ! > > Will this talk be streamed ? yes > > See you there! > The Tea Talk Team > > TITLE Machine Learning and Medicine: How AI will change the way we see patients, and the way we see ourselves. > > ABSTRACT > Just as astronomy constitutes the exploration of outer space, advances in medicine may be seen as an ever-deeper exploration of inner space. This talk will explore how that inwardly-focused exploration may be transformed in the age of machine learning. I will begin by attempting to convey a view of artificial intelligence from the vantage point of medicine: what physicians tend to feel about AI, what they tend to know, and what they generally do not know. I will briefly cite examples of productive (and less productive) emerging uses of AI in medicine. I will then focus on medical imaging in particular, and will summarize some of the goals, early outcomes, challenges, and future directions of the fastMRI collaboration between NYU School of Medicine and Facebook AI Research, in which deep learning is being used to accelerate MRI beyond previous limits. In addition to being of great value to patients, physicians, and healthcare systems, acceleration serves as a potent enabler of several emerging trends that promise to reshape the future of biomedical imaging, including a move from carefully-staged snapshots to continuous streaming, and a move from imitating the eye to emulating the brain. In this context, I will explore how AI may change not only the analysis of images and other sensor data streams, but also the design and use of future imaging devices. I will conclude with a few speculations about potential changes in the way we interact with the medical system and even how we perceive the world around us. Throughout, I will attempt to highlight areas in which data scientists can add value to the day-to-day practice of medical imaging, to the improvement of human health, and to the ongoing exploration of inner space. > > BIO > Daniel K. Sodickson MD, PhD is Vice-Chair for Research in the Department of Radiology at NYU Langone Health, Professor of Radiology and Physiology & Neuroscience at NYU School of Medicine, and Professor of Biomedical Engineering at the NYU Tandon School of Engineering. He has led a transformation of imaging research at NYU Langone, bringing the Radiology Department’s national research ranking from #17 to #5, and earning the department’s Center for Advanced Imaging Innovation and Research (CAI2R) a designation as a national Biomedical Technology Resource Center. Dr. Sodickson’s research aims at seeing what has previously been invisible, in order to improve human health. He is credited with founding the field of parallel imaging, in which distributed arrays of detectors are used to gather magnetic resonance images at previously inaccessible speeds. Parallel imaging hardware and software is now an integral part of MRI machines, and is used routinely in MRI scans worldwide. In 2006, Dr. Sodickson was awarded the Gold Medal of the International Society for Magnetic Resonance in Medicine (ISMRM), and he recently completed a term as ISMRM president. Last year, he helped to initiate the fastMRI collaboration between NYU and FAIR, and he continues to work closely with colleagues at FAIR. He is currently in the process of launching a new institute – Tech4Health – designed to bring emerging technologies such as continuous sensing and artificial intelligence to biomedicine. > > -- > You received this message because you are subscribed to the Google Groups "MILA Tous" group. > To unsubscribe from this group and stop receiving emails from it, send an email to mila-tous+unsubscribe(a)mila.quebec.

1 0

[Double Tea Talk] Marc Bellemare (Brain) x Pieter Abbeel (UC Berkeley) Fri 24 May 10h30 Mila Auditorium
by Rim Assouel 06 Mai '19

06 Mai '19

On May the 24th we will have a special double tea talk (40min each) featuring Marc G. Bellemare and Pieter Abbeel from 10.30AM to 12PM at Mila Auditorium! We send the notification email early enough so that you can all mark your calendars :) Those talks will be streamed and recorded here <https://mila.bluejeans.com/4862024040/webrtc> ! See you there, The Tea Talk Team :) TITLE Some progress on understanding the benefits of distributional reinforcement learning ABTRACT In this talk I will review what we now know about distributional reinforcement learning and how its full benefits are only obtained when combined with nonlinear representations such as deep networks. I will discuss how trying to understanding the good empirical performance of distributional RL has led us to all kinds of exciting results regarding representation learning for RL, and in particular a formulation of optimal representation learning based on the geometric notion of a value function polytope. BIO Marc G. Bellemare leads the reinforcement learning efforts at Google Brain in Montreal and holds a Canada CIFAR AI Chair at the Quebec Artificial Intelligence Institute (Mila). He received his Ph.D. from the University of Alberta, where he developed the highly-successful Arcade Learning Environment, the platform that sparked the recent revival in deep reinforcement learning research. He joined DeepMind in 2013 prior to its acquisition by Google and was research scientist there until his return to Canada in 2017. During his tenure at DeepMind he made important contributions to deep reinforcement learning, in particular pioneering the distributional method. Marc G. Bellemare is also a CIFAR Learning in Machines & Brains Fellow and an adjunct professor at McGill University. …………. TITLE Model-based RL via Meta-Model-Free RL ABSTRACT Model-free RL has seen great asymptotic successes, but sample complexity tends to be high. Model-based RL carries the promise of better sample efficiency, and indeed has shown more data efficient learning, but tends to fall well short of model-free RL in terms of asymptotic performance. In this presentation I will describe a new approach to model-based RL that brings in ideas from domain randomization and meta-model-free RL, resulting in the best of both worlds: fast learning and great asymptotic performance. Our method is evaluated on several mujoco environments (PR2 reacher, swimmer, hopper, ant, swimmer, walker) and is able to learn lego-block placement on a real robot in 10 minutes. BIO Professor Pieter Abbeel is Director of the Berkeley Robot Learning Lab and Co-Director of the Berkeley Artificial Intelligence (BAIR) Lab. Abbeel’s research strives to build ever more intelligent systems, which has his lab push the frontiers of deep reinforcement learning, deep imitation learning, deep unsupervised learning, transfer learning, meta-learning, and learning to learn, as well as study the influence of AI on society. His lab also investigates how AI could advance other science and engineering disciplines. Abbeel has founded three companies: Gradescope, Covariant, and Berkeley Open Arms, advises many AI and robotics start-ups, and is a frequently sought after speaker worldwide for C-suite sessions on AI future and strategy. Abbeel has received many awards and honors, including the PECASE, NSF-CAREER, ONR-YIP, Darpa-YFA, TR35. His work is frequently featured in the press, including the New York Times, Wall Street Journal, BBC, Rolling Stone, Wired, and Tech Review.

1 0

[Tea Talk] ICLR Lightning talks (Mila) Fri 3 May 2019 10h30 Mila Auditorium
by Rim Assouel 03 Mai '19

03 Mai '19

This week we have ICLR Lightning talks at 10h30 in the Mila Auditorium. Each signed-up presenter will give a short 5-10 min presentation of his/her accepted work at ICLR ! Will this talk be streamed <https://mila.bluejeans.com/809027115/webrtc>? yes See you there! The Tea Talk Team

2 1

Re: [Lisa_teatalk] [Tea Talk] Yoshua Bengio (Mila) Fri 12 April 2019 10h30 Mila Auditorium
by Rim Assouel 26 Avr '19

26 Avr '19

Yes, you can find the recording on the Mila youtube channel or the tea talks website! Cheers, The Tea Talk Team Le mercredi 24 avril 2019, Rémi LP <remi.lp.17(a)gmail.com> a écrit : > Was this talk recorded ? > > On Apr 12, 2019, at 09:59, Rim Assouel <rim.assouel(a)gmail.com> wrote: > > Reminder that this happens in 30 minutes :) > > Le lundi 8 avril 2019, Rim Assouel <rim.assouel(a)gmail.com> a écrit : > >> This week we have * Yoshua Bengio* from * Mila * giving a talk on * >> Meta-transfer learning for factorizing representations and knowledge for AI* >> at * 10h30* in room * Mila Auditorium*. >> >> Will this talk be streamed >> <https://mila.bluejeans.com/4862024040/webrtc>? yes >> >> See you there! >> The Tea Talk Team >> >> *TITLE* Meta-transfer learning for factorizing representations and >> knowledge for AI >> >> *ABSTRACT* >> Whereas machine learning theory has focused on generalization to examples >> from the same distribution as the training data, better understanding of >> the transfer scenarios where the observed distribution changes often in the >> lifetime of the learning agent is important, both for robust deployment and >> to achieve a more powerful form of generalization which humans seem able to >> enjoy and which seem necessary for learning agents. Whereas most machine >> learning algorithms and architectures can be traced back to assumptions >> about the training distributions, we also need to explore assumptions about >> how the observed distribution changes. We propose that sparsity of change >> in distribution, when knowledge is represented appropriately, is a good >> assumption for this purpose, and we claim that if that assumption is >> verified and knowledge represented appropriately, it leads to fast >> adaptation to changes in distribution, and thus that the speed of >> adaptation to changes in distribution can be used as a meta-objective which >> can drive the discovery of knowledge representation compatible with that >> assumption. We illustrate these ideas in causal discovery: is some variable >> a direct cause of another? and how to map raw data to a representation >> space where different dimensions correspond to causal variables for which a >> clear causal relationship exists? We propose a large research program in >> which this non-stationarity assumption and meta-transfer objective is >> combined with other closely related assumptions about the world embodied in >> a world model, such as the consciousness prior (the causal graph is >> captured by a sparse factor graph) and the assumption that the causal >> variables are often those agents can act upon (the independently >> controllable factors prior), both of which should be useful for agents >> which plan, imagine and try to find explanations for what they observe. >> >> *BIO* >> Yoshua Bengio is Full Professor in the computer science and operations >> research department at U. Montreal, scientific director of Mila and of >> IVADO, Turing Award 2018 recipient, Canada Research Chair in Statistical >> Learning Algorithms, as well as a Canada AI CIFAR Chair. He pioneered deep >> learning and has been getting the most citations per day in 2018 among all >> computer scientists, worldwide. He is officer of the Order of Canada, >> member of the Royal Society of Canada, was awarded the Marie-Victorin Prize >> and the Radio-Canada Scientist of the year in 2017, and he is a member of >> the NeurIPS board and co-founder and general chair for the ICLR conference, >> as well as program director of the CIFAR program on Learning in Machines >> and Brains. His goal is to contribute to uncover the principles giving rise >> to intelligence through learning, as well as favour the development of AI >> for the benefit of all. >> > > -- > You received this message because you are subscribed to the Google Groups > "MILA Tous" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to mila-tous+unsubscribe(a)mila.quebec. > > >

2 2

information sur les teatalk
by Stéphanie Giroux 22 Avr '19

22 Avr '19

Bonjour, Je fais présentement une maitrise en analyse de données et intelligence d’affaire au HEC. Je suis intéressée par les teatalk. Quel est le principe? Ça ressemble à une conférence ou plus à une discussion? C’est toujours en anglais? Merci de l’information. Stéphanie [https://ipmcdn.avast.com/images/icons/icon-envelope-tick-round-orange-anima…]<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campai…> Virus-free. www.avast.com<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campai…>

1 0

No Tea Talk this week
by Rim Assouel 22 Avr '19

22 Avr '19

Hi everyone, There is no tea talk this week so that you can enjoy TechAide conference :) Cheers, The Tea Talk Team

1 0

[teatalk-orgs] [Tea Talk] Dhruva Raman (Cambridge) Fri 19 April 2019 10h30 Mila Auditorium
by saikrishna gottipati 19 Avr '19

19 Avr '19

This week we have *Dhruva Raman* from *Univ of Cambridge *giving a talk on *Fundamental Bounds on Learning Performance in Neural Circuits *at *10h30* on *19th April* in *Mila Auditorium*. If you are interested in* meeting Dhruva*, please book a slot in this google sheet <https://docs.google.com/spreadsheets/d/13k6e1rCDM1tD__MvxS-UDHMS9L6XsWcTvqs…> See you there! The Tea Talk Team *TITLE* Fundamental Bounds on Learning Performance in Neural Circuits *ABSTRACT* Biological neural circuits learn in spite of imperfect information on task performance and noisy biological components. How can these problems be mitigated? We use optimization theory to show how adding apparently redundant neurons and connections to a network can improve learning performance in the face of imperfect learning rules and corrupted error signals. The theory shows how large neural circuits can exploit additional connectivity to achieve faster and more precise learning. However, there is a limit to the benefit of adding connections. Biologically, synapses (connections strengths) are intrinsically unreliable. We show that excessive network size eventually outcompetes the benefits to learning performance. Consequently, there is an optimal size of network for a given task, which we can calculate in specific cases. *BIO* Dhruva did his undergraduate (MMath) at the University of Warwick (2008-2012), spent a year at the systems biology doctoral training centre at the University of Oxford (2012-2013), and did his PhD in the Control Group at the University of Oxford under the supervision of Antonis Papachristodoulou (2013-2016). Since 2017 he has been a postdoc in the control group at the University of Cambridge under the supervision of Timothy O’Leary.

2 2

[Tea Talk] Yoshua Bengio (Mila) Fri 12 April 2019 10h30 Mila Auditorium
by Rim Assouel 12 Avr '19

12 Avr '19

This week we have Yoshua Bengio from Mila giving a talk on Meta-transfer learning for factorizing representations and knowledge for AI at 10h30 in room Mila Auditorium. Will this talk be streamed <https://mila.bluejeans.com/4862024040/webrtc>? yes See you there! The Tea Talk Team TITLE Meta-transfer learning for factorizing representations and knowledge for AI ABSTRACT Whereas machine learning theory has focused on generalization to examples from the same distribution as the training data, better understanding of the transfer scenarios where the observed distribution changes often in the lifetime of the learning agent is important, both for robust deployment and to achieve a more powerful form of generalization which humans seem able to enjoy and which seem necessary for learning agents. Whereas most machine learning algorithms and architectures can be traced back to assumptions about the training distributions, we also need to explore assumptions about how the observed distribution changes. We propose that sparsity of change in distribution, when knowledge is represented appropriately, is a good assumption for this purpose, and we claim that if that assumption is verified and knowledge represented appropriately, it leads to fast adaptation to changes in distribution, and thus that the speed of adaptation to changes in distribution can be used as a meta-objective which can drive the discovery of knowledge representation compatible with that assumption. We illustrate these ideas in causal discovery: is some variable a direct cause of another? and how to map raw data to a representation space where different dimensions correspond to causal variables for which a clear causal relationship exists? We propose a large research program in which this non-stationarity assumption and meta-transfer objective is combined with other closely related assumptions about the world embodied in a world model, such as the consciousness prior (the causal graph is captured by a sparse factor graph) and the assumption that the causal variables are often those agents can act upon (the independently controllable factors prior), both of which should be useful for agents which plan, imagine and try to find explanations for what they observe. BIO Yoshua Bengio is Full Professor in the computer science and operations research department at U. Montreal, scientific director of Mila and of IVADO, Turing Award 2018 recipient, Canada Research Chair in Statistical Learning Algorithms, as well as a Canada AI CIFAR Chair. He pioneered deep learning and has been getting the most citations per day in 2018 among all computer scientists, worldwide. He is officer of the Order of Canada, member of the Royal Society of Canada, was awarded the Marie-Victorin Prize and the Radio-Canada Scientist of the year in 2017, and he is a member of the NeurIPS board and co-founder and general chair for the ICLR conference, as well as program director of the CIFAR program on Learning in Machines and Brains. His goal is to contribute to uncover the principles giving rise to intelligence through learning, as well as favour the development of AI for the benefit of all.

1 1

[Tea Talk] Anirudh Goyal (Mila) Fri 5 April 2019 10h30 Mila Auditorium
by Rim Assouel 05 Avr '19

05 Avr '19

This week we have Anirudh Goyal from Mila giving a talk on Inductive Biases, Invariances and Generalization at 10h30 in room Mila Auditorium. Will this talk be streamed <https://mila.bluejeans.com/809027115/webrtc>? yes See you there! The Tea Talk Team TITLE Inductive Biases, Invariances and Generalization ABSTRACT The biggest challenge right now in RL specifically, AI more generally, is to devise methods for learning complex behaviours that have meaningful generalization. In supervised learning (vision, speech, NLP), it seems that deep models can achieve very complex generalization, but this is not really the case in RL. In this talk, I'm going to first argue that this gap in generalization has much to do with current practices in evaluating generalization in multitask, transfer and meta-learning setups, and then I'm going to talk about an elegant solution for learning complex inductive biases that can be used to exploit the structure of the task, thus enabling the learning agent to incorporate prior knowledge into the learning system to exploit reusable structure in task space. BIO Anirudh Goyal is a first year PhD student advised by Prof. Yoshua Bengio. His current research interests center around understanding the interplay b/w three words: Invariances, Inductive Biases and Generalization.

2 2

[Coffee-Talk] April Van Houtte (Mila) Tuesday 1 April 2020 6h30 Mila Auditorium
by Rim Assouel 01 Avr '19

01 Avr '19

Next year we will have an additional Coffee talk given by April Van Houtte from Mila giving a talk on Generalization of the S-o-C in Deep Learning at 6h30 in room Mila Auditorium. Will this talk be streamed ? yes See you there! Pablo, Sai and Rim for The Coffee Talk Team TITLE Biases in Grad students: Analysis from a Cafeine Theoretic perspective ABSTRACT The biggest challenge right now in grad students specifically, AI grad students, is to devise methods for learning complex behaviours that include attending classes at 8.30am. In this talk, I'm going to first argue that the discrepancy between tea and coffee has much to do with current practices in evaluating generalization in multitask, transfer and meta-learning setups, and then I'm going to talk about an elegant solution for learning complex productive working strategies that can be used to exploit the structure of the task, thus enabling the grad student to incorporate prior knowledge into the learning system to exploit reusable structure in task space. BIO April Van Houtte is a senior member of Mila and he works on keeping grad students caffeinated. He pioneered the S-o-C (sip of coffee) operator as a new neural modulation block.He recently received enduring award from ACM (Association of Coffee Makers) which is considered the noble prize in hot beverages.

1 0

[Tea Talk] Jeffrey Pennington (Google Brain NYC) Fri 29 March 2019 10h30 Mila Auditorium
by Rim Assouel 29 Mar '19

29 Mar '19

This week we have Jeffrey Pennington from Google Brain NYC giving a talk on Are Overparameterized Neural Networks Actually Just Linear Models? at 10h30 in room Mila Auditorium. Will this talk be streamed <https://mila.bluejeans.com/809027115/webrtc>? yes See you there! The Tea Talk Team TITLE Are Overparameterized Neural Networks Actually Just Linear Models? ABSTRACT Neural networks define a rich and expressive class of functions whose properties and behaviors are very hard to describe from a theoretical perspective. Nevertheless, when these functions become highly overparameterized, a surprisingly simple characterization emerges. In this talk, I will discuss several perspectives on this characterization: 1) I will examine the prior over functions induced by common weight initialization schemes and show that it corresponds to a Gaussian process with a well-defined compositional kernel; 2) I will show that by tuning initialization hyperparameters, this kernel can be optimized for signal propagation, yielding networks that are trainable to enormous depths (10k+ layers); and 3) I will demonstrate that the learning dynamics of such overparameterized neural networks are governed by a linear model obtained from the first-order Taylor expansion of the network around its initial parameters. BIO Jeffrey Pennington is a Research Scientist at Google Brain, NYC. Prior to this, he was a postdoctoral fellow at Stanford University, as a member of the Stanford Artificial Intelligence Laboratory in the Natural Language Processing (NLP) group, where he studied the unsupervised learning of word representations. He received his Ph.D. in theoretical particle physics from Stanford University while working at the SLAC National Accelerator Laboratory, with a main focus on the development of calculational techniques in perturbative quantum field theory. Jeffrey’s current research interests center on the theory of deep learning, and include topics such as: trainability and expressivity, the dynamics of learning, the role of overparameterization, stochastic networks and random matrix theory, and the geometry of high-dimensional loss surfaces

2 2

[mila-tous] [Discussion group + Tea Talk] David Blei (Columbia University) Fri March 8 2019 10h30 Mila Auditorium
by Rim Assouel 06 Mar '19

06 Mar '19

Hi everyone, Reminder that David arrives tomorrow and we’ll be having the Discussion group from 2pm to 3pm in room A14 : everybody is welcome to join! Do not miss his tea talk on Friday, 10.30 in Mila auditorium as well :) Cheers, The Tea Talk Team > On Mar 5, 2019, at 4:59 AM, Yoshua Bengio <yoshua.bengio(a)mila.quebec> wrote: > > I concur! Do not miss David's talk and come to meet him too. > > -- Yoshua > > SVP prendre note de mon nouveau courriel ainsi que de celui de mon adjointe : > Please note my new email address as well as my assistant's one: > yoshua.bengio(a)mila.quebec > julie.mongeau(a)mila.quebec > > > Le lun. 4 mars 2019, à 12 h 15, Laurent Charlin <lcharlin(a)gmail.com <mailto:lcharlin@gmail.com>> a écrit : > Dave has done foundational work and is a leading thinker in the fields of probabilistic graphical models and variational inference. He is also an excellent speaker. His talk should not be missed! > > Best, > Laurent > > On Mon, Mar 4, 2019 at 10:30 AM Rim Assouel <rim.assouel(a)gmail.com <mailto:rim.assouel@gmail.com>> wrote: > This week we have David Blei from Columbia University giving a talk on The Blessings of Multiple Causes at 10h30 at Mila Auditorium. > > Will this talk be streamed <https://mila.bluejeans.com/809027115/webrtc>? Yes > > David will also be visiting Mila on Thursday (7th March). To that end we organized a discussion group on Causal Inference from 2pm to 3pm in room A14. > > A bunch of you asked for 1-1 and 2-1 meetings : The schedule is also set during that day, I’ll send the invites by tomorrow :) > > See you there! > The Tea Talk Team > > TITLE The Blessings of Multiple Causes > > ABSTRACT > Causal inference from observational data is a vital problem, but it comes with strong assumptions. Most methods require that we observe all confounders, variables that correlate to both the causal variables (the treatment) and the effect of those variables (how well the treatment works). But whether we have observed all confounders is a famously untestable assumption. We describe the deconfounder, a way to do causal inference from observational data with weaker assumptions that the classical methods require. How does the deconfounder work? While traditional causal methods measure the effect of a single cause on an outcome, many modern scientific studies involve multiple causes, different variables whose effects are simultaneously of interest. The deconfounder uses the multiple causes as a signal for unobserved confounders, combining unsupervised machine learning and predictive model checking to perform causal inference. We describe the theoretical requirements for the deconfounder to provide unbiased causal estimates, and show that it requires weaker assumptions than classical causal inference. We analyze the deconfounder's performance in three types of studies: semi-simulated data around smoking and lung cancer, semi-simulated data around genomewide association studies, and a real dataset about actors and movie revenue. The deconfounder provides a checkable approach to estimating close-to-truth causal effects. This is joint work with Yixin Wang. [*] https://arxiv.org/abs/1805.06826 <https://arxiv.org/abs/1805.06826> > > BIO > David Blei is a Professor of Statistics and Computer Science at Columbia University, and a member of the Columbia Data Science Institute. He studies probabilistic machine learning, including its theory, algorithms, and application. David has received several awards for his research, including a Sloan Fellowship (2010), Office of Naval Research Young Investigator Award (2011), Presidential Early Career Award for Scientists and Engineers (2011), Blavatnik Faculty Award (2013), ACM-Infosys Foundation Award (2013), and a Guggenheim fellowship (2017). He is the co-editor-in-chief of the Journal of Machine Learning Research. He is a fellow of the ACM and the IMS. He is a fellow of the ACM and the IMS. > > -- > You received this message because you are subscribed to the Google Groups "MILA Tous" group. > To unsubscribe from this group and stop receiving emails from it, send an email to mila-tous+unsubscribe(a)mila.quebec <mailto:mila-tous+unsubscribe@mila.quebec>. > > -- > You received this message because you are subscribed to the Google Groups "MILA Tous" group. > To unsubscribe from this group and stop receiving emails from it, send an email to mila-tous+unsubscribe(a)mila.quebec <mailto:mila-tous+unsubscribe@mila.quebec>.

1 0

[Tea Talk] David Blei (Columbia University) Fri March 8 2019 10h30 Mila Auditorium
by Rim Assouel 04 Mar '19

04 Mar '19

This week we have David Blei from Columbia University giving a talk on The Blessings of Multiple Causes at 10h30 at Mila Auditorium. Will this talk be streamed <https://mila.bluejeans.com/809027115/webrtc>? Yes David will also be visiting Mila on Thursday (7th March). To that end we organized a discussion group on Causal Inference from 2pm to 3pm in room A14. A bunch of you asked for 1-1 and 2-1 meetings : The schedule is also set during that day, I’ll send the invites by tomorrow :) See you there! The Tea Talk Team TITLE The Blessings of Multiple Causes ABSTRACT Causal inference from observational data is a vital problem, but it comes with strong assumptions. Most methods require that we observe all confounders, variables that correlate to both the causal variables (the treatment) and the effect of those variables (how well the treatment works). But whether we have observed all confounders is a famously untestable assumption. We describe the deconfounder, a way to do causal inference from observational data with weaker assumptions that the classical methods require. How does the deconfounder work? While traditional causal methods measure the effect of a single cause on an outcome, many modern scientific studies involve multiple causes, different variables whose effects are simultaneously of interest. The deconfounder uses the multiple causes as a signal for unobserved confounders, combining unsupervised machine learning and predictive model checking to perform causal inference. We describe the theoretical requirements for the deconfounder to provide unbiased causal estimates, and show that it requires weaker assumptions than classical causal inference. We analyze the deconfounder's performance in three types of studies: semi-simulated data around smoking and lung cancer, semi-simulated data around genomewide association studies, and a real dataset about actors and movie revenue. The deconfounder provides a checkable approach to estimating close-to-truth causal effects. This is joint work with Yixin Wang. [*] https://arxiv.org/abs/1805.06826 BIO David Blei is a Professor of Statistics and Computer Science at Columbia University, and a member of the Columbia Data Science Institute. He studies probabilistic machine learning, including its theory, algorithms, and application. David has received several awards for his research, including a Sloan Fellowship (2010), Office of Naval Research Young Investigator Award (2011), Presidential Early Career Award for Scientists and Engineers (2011), Blavatnik Faculty Award (2013), ACM-Infosys Foundation Award (2013), and a Guggenheim fellowship (2017). He is the co-editor-in-chief of the Journal of Machine Learning Research. He is a fellow of the ACM and the IMS. He is a fellow of the ACM and the IMS.

2 1

[NO Tea Talk this week]
by Rim Assouel 07 Fév '19

07 Fév '19

Hi everyone, To let people attend the scheduled job talk, there will be no tea talk this week. Sai, Pablo and Rim

1 0

[Tea Talk] Isabela Albuquerque (INRS) Fri February 1 2019 10h30 Mila Auditorium
by Rim Assouel 01 Fév '19

01 Fév '19

This week we have Isabela Albuquerque from INRS giving a talk on Multi-objective training of Generative Adversarial Networks with multiple discriminators at 10h30 in room Mila Auditorium. Will this talk be streamed <https://mila.bluejeans.com/809027115/webrtc>? yes PS : Slides and recordings of previous talks are available at :https://sites.google.com/lisa.iro.umontreal.ca/tea-talk-recordings/home?authuser=0 <https://sites.google.com/lisa.iro.umontreal.ca/tea-talk-recordings/home?aut…> See you there! Pablo, Sai and Rim TITLE Multi-objective training of Generative Adversarial Networks with multiple discriminators ABSTRACT Recent literature has demonstrated promising results for training Generative Adversarial Networks by employing a set of discriminators, in contrast to the traditional game involving one generator against a single adversary. Such methods perform single-objective optimization on some simple consolidation of the losses, e.g. an average. In this work, we revisit the multiple-discriminator setting by framing the simultaneous minimization of losses provided by different models as a multi-objective optimization problem. Specifically, we evaluate the performance of multiple gradient descent and the hypervolume maximization algorithm on a number of different datasets. Moreover, we argue that the previously proposed methods and hypervolume maximization can all be seen as variations of multiple gradient descent in which the update direction can be computed efficiently. Our results indicate that hypervolume maximization presents a better compromise between sample quality and computational cost than previous methods. BIO Isabela Albuquerque is a PhD candidate at the Institut National de la Recherche Scientifique (INRS), Université du Québec. Her main research interests are machine learning and cognitive neuroscience. The primary goal of her current research is improving the generalization across different individuals of electroencephalogram-based mental workload models. Prior to joining INRS, she received a Bachelor's degree in Electrical Engineering from the Federal University of Pernambuco and a Master's degree in Computer Engineering from the University of Pernambuco, both in Recife, Brazil.

2 2

[Tea Talk] Hugo Larochelle (Mila + Brain) Fri January 25 2019 10h30 Mila Auditorium
by Rim Assouel 25 Jan '19

25 Jan '19

This week we have Hugo Larochelle from Mila + Brain giving a talk on (Part II) Few-Shot Learning with Meta-Learning: Progress Made and Challenges Ahead at 10h30 in room Mila Auditorium. Will this talk be streamed <https://mila.bluejeans.com/809027115/webrtc>? yes Feel like learning to learn again ? Hugo will teach you another way, this Friday, one shot at a time ;) See you there! Pablo, Sai and Rim Tea Talks website : https://sites.google.com/lisa.iro.umontreal.ca/tea-talk-recordings <https://sites.google.com/lisa.iro.umontreal.ca/tea-talk-recordings> TITLE (Part II) Few-Shot Learning with Meta-Learning: Progress Made and Challenges Ahead ABSTRACT A lot of the recent progress on many AI tasks was enable in part by the availability of large quantities of labeled data. Yet, humans are able to learn concepts from as little as a handful of examples. Meta-learning is a very promising framework for addressing the problem of generalizing from small amounts of data, known as few-shot learning. In meta-learning, our model is itself a learning algorithm: it takes as input a training set and outputs a classifier. For few-shot learning, it is (meta-)trained directly to produce classifiers with good generalization performance for problems with very little labeled data. In this talk, I'll present an overview of the recent research that has made exciting progress on this topic (including my own) and will discuss the challenges as well as research opportunities that remain. BIO Hugo Larochelle is Research Scientist at Google Brain and lead of the Montreal Google Brain team. He is also a member of Yoshua Bengio's Mila and an Adjunct Professor at the Université de Montréal. Previously, he was Associate Professor at the University of Sherbrooke. He also co-founded Whetlab, which was acquired in 2015 by Twitter, where he then worked as a Research Scientist in the Twitter Cortex group. From 2009 to 2011, he was also a member of the machine learning group at the University of Toronto, as a postdoctoral fellow under the supervision of Geoffrey Hinton. He obtained his Ph.D. at the Université de Montréal, under the supervision of Yoshua Bengio. He has the best hair in machine learning academia (not his words). Finally, he has a popular online course on deep learning and neural networks, freely accessible on YouTube.

2 1

[No Tea Talk][ICML deadline]
by Rim Assouel 14 Jan '19

14 Jan '19

There will be no Tea Talk this week because of the ICML deadline coming soon. Next talk will be on the 25th for Hugo Larochelle’s part 2 on Few-Shot Learning with Meta-Learning. See you next week, Pablo, Sai and Rim

1 0

[Tea Talk] [New Website]Maxime Chevalier-Boisvert (Mila) Fri January 11 2019 10h30 Mila Auditorium
by Rim Assouel 11 Jan '19

11 Jan '19

This week we have Maxime Chevalier-Boisvert from Mila giving a talk on BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learning at 10h30 in room Mila Auditorium. Will this talk be streamed <https://mila.bluejeans.com/809027115/webrtc>? yes We are also pleased to share the new « website » where we will keep track of the talks slides/recordings : https://sites.google.com/lisa.iro.umontreal.ca/tea-talk-recordings/home?aut… <https://sites.google.com/lisa.iro.umontreal.ca/tea-talk-recordings/home?aut…> It’s accessible with your Mila google account :) See you there and Happy New Year! Pablo, Sai and Rim TITLE BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learning ABSTRACT This talk will present BabyAI, a research project based at Mila with the long-term goal of creating agents which we can communicate with and teach new concepts using natural language. We will begin by discussing why language learning is a hard problem, one to which deep learning doesn't yet have a satisfying answer for. We will then introduce the BabyAI platform, created to study the sample efficiency of grounded language learning in the context of embodied agents and instruction following (see paper accepted at ICLR19, https://openreview.net/forum?id=rJeXCo0cYX) Lastly, we will discuss multiple promising research directions we have identified with the goal of improving the sample efficiency of grounded language learning. BIO Maxime has completed a PhD in computer science at the Université de Montréal in early 2016. She has since worked at Apple as part of the mobile GPU team, and then joined Mila as part of the research engineering staff. At Mila, she has been co-directing the BabyAI project with Dzmitry Bahdanau, and took part in creating the Duckietown Simulator. Her main research interests are robotics, embodied agents and reinforcement learning.

3 3

Re: [Lisa_teatalk] [TODAY][Tea Talk] Praneeth (MSR India) Thursday December 13 2018 15:30 AA3195
by Rim Assouel 13 Déc '18

13 Déc '18

Reminder : this happens in 1h ! > Le 11 déc. 2018 à 17:20, rim.assouel(a)gmail.com a écrit : > > This week we have Praneeth from MSR India giving a talk on Thursday December 13 2018 at 15:30 in room AA3195 > > Will this talk be streamed <https://mila.bluejeans.com/4255239897/webrtc>? Yes > > Pay attention to the day change !! This tea talk will happen on THURSDAY :) > > As it will be the last tea talk of the year, it will be featured with actual tea, talk and snacks :) > > See you there! > Rim and Sai > > TITLE On momentum methods and acceleration in stochastic optimization > > ABSTRACT > > It is well known that momentum gradient methods (e.g., Polyak's heavy ball, Nesterov's acceleration) yield significant improvements over vanilla gradient descent in deterministic optimization (i.e., where we have access to exact gradient of the function to be minimized). However, there is widespread sentiment that these momentum methods are not effective for the purposes of stochastic optimization due to their instability and error accumulation. Numerous works have attempted to quantify these instabilities in the face of either statistical or non-statistical errors (Paige, 1971; Proakis, 1974; Polyak, 1987; Greenbaum, 1989; Roy and Shynk, 1990; Sharma et al., 1998; d’Aspremont, 2008; Devolder et al., 2013, 2014; Yuan et al., 2016) but a precise understanding is lacking. This work considers these issues for the special case of stochastic approximation for the linear least squares regression problem, and shows that: > > 1. classical momentum methods (heavy ball and Nesterov's acceleration) indeed do not offer any improvement over stochastic gradient descent, and > 2. introduces an accelerated stochatic gradient method that provably achieves the minimax optimal statistical risk faster than stochastic gradient descent (and classical momentum methods). > > Critical to the analysis is a sharp characterization of accelerated stochastic gradient descent as a stochastic process. While the results are rigorously established for the special case of linear least squares regression, experiments suggest that the conclusions hold for the training of deep neural networks. > BIO > Praneeth Netrapalli is a researcher at Microsoft Research India, Bengaluru since August 2016. Prior to this, he was a postdoctoral researcher at Microsoft Research New England in Cambridge, MA. He obtained MS and PhD from UT Austin and B-Tech from IIT Bombay all in Electrical Engineering. His research focuses on designing efficient algorithms for machine learning problems primarily via stochastic and nonconvex optimization. More information about his research is available on his home page http://praneethnetrapalli.org/

2 1

[DAY CHANGE][Tea Talk] Praneeth (MSR India) Thursday December 14 2018 15:30 AA3195
by rim.assouel＠gmail.com 12 Déc '18

12 Déc '18

This week we have Praneeth from MSR India giving a talk on Thursday December 13 2018 at 15:30 in room AA3195 Will this talk be streamed <https://mila.bluejeans.com/4255239897/webrtc>? Yes Pay attention to the day change !! This tea talk will happen on THURSDAY :) As it will be the last tea talk of the year, it will be featured with actual tea, talk and snacks :) See you there! Rim and Sai TITLE On momentum methods and acceleration in stochastic optimization ABSTRACT It is well known that momentum gradient methods (e.g., Polyak's heavy ball, Nesterov's acceleration) yield significant improvements over vanilla gradient descent in deterministic optimization (i.e., where we have access to exact gradient of the function to be minimized). However, there is widespread sentiment that these momentum methods are not effective for the purposes of stochastic optimization due to their instability and error accumulation. Numerous works have attempted to quantify these instabilities in the face of either statistical or non-statistical errors (Paige, 1971; Proakis, 1974; Polyak, 1987; Greenbaum, 1989; Roy and Shynk, 1990; Sharma et al., 1998; d’Aspremont, 2008; Devolder et al., 2013, 2014; Yuan et al., 2016) but a precise understanding is lacking. This work considers these issues for the special case of stochastic approximation for the linear least squares regression problem, and shows that: 1. classical momentum methods (heavy ball and Nesterov's acceleration) indeed do not offer any improvement over stochastic gradient descent, and 2. introduces an accelerated stochatic gradient method that provably achieves the minimax optimal statistical risk faster than stochastic gradient descent (and classical momentum methods). Critical to the analysis is a sharp characterization of accelerated stochastic gradient descent as a stochastic process. While the results are rigorously established for the special case of linear least squares regression, experiments suggest that the conclusions hold for the training of deep neural networks. BIO Praneeth Netrapalli is a researcher at Microsoft Research India, Bengaluru since August 2016. Prior to this, he was a postdoctoral researcher at Microsoft Research New England in Cambridge, MA. He obtained MS and PhD from UT Austin and B-Tech from IIT Bombay all in Electrical Engineering. His research focuses on designing efficient algorithms for machine learning problems primarily via stochastic and nonconvex optimization. More information about his research is available on his home page http://praneethnetrapalli.org/

3 2

[No tea-talk][NeurIPS][Predocs]
by Rim Assouel 06 Déc '18

06 Déc '18

There will not be any tea talk tomorrow due to NeurIPS! See you next week, PS : Tong Che’s predoc III on Wednesday at 1pm PPS : Gabriel Huang’s predoc III on Wednesday at 10 am Rim & Sai

1 0

[Tea Talk] NeurIPS accepted papers (Mila) Fri November 30 2018 10:30 AA5340
by rim.assouel＠gmail.com 30 Nov '18

30 Nov '18

This week we have NeurIPS accepted papers from Mila giving a talk on Fri November 30 2018 at 10:30 in room AA5340 TITLE Happy NeurIPS :) ABSTRACT Carrying on the fantastic tradition, we are organizing another lightning talk, this time for NeurIPS! We’ll have authors of NeurIPS accepted papers do a quick, ~5 minute presentations of their work. Will this talk be streamed <https://mila.bluejeans.com/4255239897/webrtc>? Yes See you there! Rim and Sai

3 2

Question
by Hadi 28 Nov '18

28 Nov '18

Hello, I wonder if I request for the information in the training section of the web-site. Is the course list (including professors) for the winter 2018 the same for winter 2019? Sincerely yours, -- *Hadi Abdi Ghavidel* *MSc. Graduate in Computational Linguistics, * *Sharif University of Technology * *PhD Student in Computer Engineering (Polytechnique), * *Université de Montréal*

2 1

MIML talk, Matt Jacobs (UCLA) 11am Tuesday: acceleration of primal-dual algorithm with application to Wasserstein distance
by Adam Oberman 26 Nov '18

26 Nov '18

Hi, Please post the following talk, which may be of interest to the Lisa tea talk group: Matt Jacobs from UCLA will speak Tuesday (tomorrow) at 11:00am in Burnside 1214. (Burnside is the McGill mathematics building, located near the corner of University and Sherbrooke). https://dms.umontreal.ca/~mathapp/machinelearning/MattJacobs.html <https://dms.umontreal.ca/~mathapp/machinelearning/MattJacobs.html> He is speaking on a faster optimization algorithm applied to total variation denoising and the Wasserstein distance Please join for a (non-hosted) lunch after the seminar. Adam Adam Oberman | Professor | Department of Mathematics | McGill University | oberman.adam(a)gmail.com | http://www.adamoberman.net/

1 0