Gail Murphy

I'm excited that this week at the CAIAC - Canadian Artificial Intelligence Association conference George Wang will be presenting a paper Adaptive Learning Rates for Gradient Boosting Trees, based on work he did as an intern at Kinaxis last fall, in collaboration with Christopher Wang, Yunfei O., Behrouz Soleimani, PhD. There are 3 reasons this is both impressive and cool: 1. They combine heuristics, machine learning, and optimization to speed up the training for an important algorithm, Gradient Boosted Machines, widely used in both academia and industry. Faster training improves model performance. 2. This paper is just one of the many sparks of innovation in #artificialintelligence coming out of Kinaxis, where > 50% of our patents filed in the last several years have been in this space, because we continue to push the boundaries by inventing new solutions like this one to solve problems our customers face. We've been delivering AI to the market since 2018 and were the first to provide solutions powered by AI to address both demand and supply. We are rooted in Canada, which brought the world out of the last AI winter in 2012 when a team from the University of Toronto won the famous Imagenet competition by blowing the lid off previous results. Their submission was layers of neural networks called deep learning, which later became the foundation for the #generativeAI revolution we are seeing today. 3. We have an outstanding intern program in #machinelearning at Kinaxis, where interns work on real problems, guided by senior staff. As a testament to letting the students shine, when their paper was accepted they decided to have George present it. So check out the paper if you're at the conference, but if you miss the event you can't miss the fact that we are actively speaking and sparking on AI. In fact, just last week Chantal Bisson-Krol spoke in Edmonton at the #upperbound conference on AI-Powered Strategies for Resilient Supply Chains and I spoke at the Reuters Supply Chain USA event on Embedding AI in supply chain to deliver end-to-end value.

43

2 Comments

This is so exciting. For the Nobel body, for the scientific community, and for the world to acknowledge the tremendous work done to enable AI and ML as we know it today. The shift in paradigm that was created with Jeffrey's work (and the co-recipient's) has made work in this space, deep, philosophical and applicable at same time. https://lnkd.in/ggW-q45K #machinelearning #ai #artificialintelligence Congratulations form all of us at Mesh AI (collaborative scheduling to end clinician burnout).

36

2 Comments

The Special Issue on Graph Learning of IEEE Transactions on Neural Networks and Learning Systems is now officially published! Guest edited by Feng Xia, Renaud Lambiotte, Neil Shah, Hanghang Tong, Irwin King. This issue features groundbreaking research that pushes the boundaries of graph learning, neural networks, and their applications across diverse fields. I extend my gratitude to all the contributors, reviewers, and the editorial team who made this possible. If you are interested in the latest developments in graph learning, this collection is a must-read! #GraphLearning #AI #MachineLearning #NeuralNetworks #IEEE #SpecialIssue

57

Today’s Nobel Prize in Physics Congratulations to Geoffrey Hinton for being a co-recipient of the 2024 Nobel Prize in physics for his work in Machine Learning. Some articles this morning have coined him as the grandfather of AI.   Dr. Hinton is a favourite topic of Canadian IP pundits. By all accounts his AI work at UoT was groundbreaking and, along the way, patents associated with some of it are now owned by Google. It is here that the pundits entered the picture with chatter about the loss of Canadian IP.   I too entered the chatter in 2020, sort of. Okay, it was more like talking to myself. At that time I penned a series of three articles about IP in Canada. The first or these (https://lnkd.in/esE7eKc) considered the loss of technology and IP. I said we need to identify the technology that is being developed and patented and we need to invest.  We need to be on the same playing field as those buying Canadian IP. This is one of my motivations for founding IPmart.   A September 15th Opinion in the Globe & Mail framed the IP discussion in terms of productivity. The authors wrote “… we need to dramatically increase the rate at which we protect and commercialize new ideas and the intellectual property (IP) that often starts in publicly funded research.” Hmmm … that sounds familiar! #technology #patents #CDNtech #IPmart

7

One of my favourite podcasts in the past few months. Peter Attia is both a medical doctor and and engineer. In this podcast, he discusses AI with Isaac Kohane, a medical doctor with a PhD in Computer Science. The first half they do a great job of exploring the history of AI and how we got today. The second half dives into AI in medicine. There is huge potential to boost medical care with further application of AI that exists today. https://lnkd.in/gZK3tcHc

6

Very pleased to report that my article with Hunter Gentry, "Transitional gradation and the distinction between episodic and semantic memory", has just appeared in _Philosophical Transactions of the Royal Society B". It is also open access thanks to the new resources I have available at University of Florida, specifically an OA agreement with the UF library and the Royal Society. So everybody can see it without any shenanigans! In this article we explore various arguments against Tulving's traditional episodic/semantic memory distinction, provide them on a firmer metaphysical foundation by drawing upon the theory of natural kinds and previous things I've said about a phenomenon called "transitional gradation" in the special sciences, recommend replacing the distinction with a more dynamic life cycle of mnemonic forms of representation, and sketch some new ideas for experiments inspired by the perspective that can help arbitrate the debate over episodic memory in animals (largely Hunter's work here). Hunter was a joy to work with on this and if you see an application from him you should give it serious consideration. He's the real deal and will be publishing stuff like this for decades to come. https://lnkd.in/gPMJSV7Q

14

Elon Musk's Neuralink announce Data Compression Challenge open to public.  The main idea of the invention is to provide efficient interface between brain and machine/computer. Implant in the brain is expected to transfer wirelessly huge amounts of data to computer and, possibly, back. It's clear that the system has many restrictions like: limited energy reserve and limited computational power. Even if theoretically compression at the 200/1 rate would be feasible (what is not), the limitations of energy resources and computation would not allow to implement that sophisticated and computationally intensive transformation in acceptable cost.  https://lnkd.in/eS-r5DuS

5

Exciting new developments in AI research! Two new papers from computer scientist Jeff Clune's lab at the University of British Columbia introduce Intelligent Go-Explore, a system that leverages a large language model, GPT-4, to select "promising" states from an archive and make intelligent exploration decisions. This approach outperformed other methods in tasks like collecting coins in games, showcasing the power of integrating language models in AI systems. Learn more about this innovative work here: https://lnkd.in/dmNM7Y-i

1

1 Comment

Uploaded a very fun talk at ACM HOPE 2024 on some new work with Cyrus Omar and Patrick Ferris on how we can formally specify systems to be robust to code generation by AI agents. For instance, if you were to ask GitHub Copilot to generate you code to filter endangered animals out of a folder of images, it might interpret that as to delete the image, or to move it to another folder (which might be public), or just remove it from the index. Any of those options are potentially valid, so what do we do? Our idea is to use F* to specify a rich set of allowable behaviours which can then be dynamically enforced in less expressive languages, and thus offer layers of protection against over-eager (or rogue) AI agents. We'll increasingly need these sort of protections in our systems as natural language interfaces get adopted for programming, and the traditional 'yes or not' access control policies we use right now will be insufficient. https://lnkd.in/e7ubvbsg for the paper and the slides at https://lnkd.in/e44iXU8h walk through the story.

8

Are you thinking about participating in the Quantum Resource Estimation challenge, but don't know where to start? Check out the video tutorial on using Azure Quantum Resource Estimator! https://lnkd.in/gqGUhzp5

50

1 Comment

New Research in Progress Monitoring! Dependable Tools for Monitoring Academic Enablers in K-5 Check out the latest study by Jobi Tat Shing Yeung, Volpe Robert, Amy Briesch, Brian Daniels, and myself on Direct Behavior Rating Multi-Item Scales (DBR-MIS). This research explores the reliability of DBR-MIS for assessing key academic enablers like academic engagement, interpersonal skills, and study skills in young learners. Highlights include: ✅ High dependability: Even with just 1-4 assessment occasions, these tools provide accurate data for progress monitoring. ✅ Individualized and Flexible: Teachers can tailor scales to the specific needs of each student, with up to four customizable items. For educators, school psychologists, and those in the field of educational assessment, this tool is very helpful for school-based applications. Dive into the full article published in the Journal of School Psychology with free access until December 22, 2024! 🔗https://lnkd.in/dEjMEXDw #Education #BehaviorRating #AcademicSuccess #SchoolPsychology #ProgressMonitoring

14

1 Comment

I will be hosting Joseph Jay Williams (Experimentation)'s visit to UBC, where he will present his talk on "Changing Behavior using AI and Experimentation" on Friday, September 20th. Please check the details and join us for what promises to be an engaging and insightful discussion! Title: Changing Behaviour using AI and Experimentation: Applications in Education, Health Behaviour, Marketing, Website Design Speaker: Joseph Jay Williams, Assistant Professor, Computer Science, University of Toronto Time and Date: 10 - 11:30 AM, September 20th, 2024 Room: ICCS 246 Zoom: https://lnkd.in/gwxSaZcM Abstract:How can we change people's behaviour to help them achieve their goals? This lecture presents tools and methods for optimizing a product or intervention by making it Intelligent and Adaptive, including: (1) Messages/Prompts in Mobile Apps; (2) Explanations of concepts in websites; (3) Texts & Emails for Marketing & Health. This work is based on the AdapEx framework,which recently received 1st prize in the $1M Xprize competition for future of experimentation, to transform components of an interface into Intelligent agents, that integrate human intelligence (crowdsourcing) and artificial intelligence (LLMs). These components become Adaptive by using perpetually MicroExperiments to test out these different actions, and using qualitative and quantitative data to optimize what to show to a user/customer, in different contexts. We show how to use AdapComps to integrate AI/LLMs, Machine Learning for adaptive experiments, theories of psychology, and human-computer interaction/user experience design.

17

3 Comments

Dear Respected Author, Upon the authors' request, the deadline for paper submission in the 2024 IEEE International Conference on Smart Mobility has been extended until May 31, 2024! The 2024 IEEE International Conference on Smart Mobility (SM), co-organized by the IEEE TEMS and Ontario Tech University, invites researchers, academics, practitioners, and industry experts to submit original research papers and contribute to third edition of the conference held under the theme "Towards People-Centric, Software-defined, Connected, and Electric Mobility". The 2024 IEEE International Conference on Smart Mobility will be held from the 16th to the 18th of September, 2024 at Niagara Falls, Ontario, Canada. The conference is designated for reporting the latest research, development, and deployment results in smart mobility systems and services, their challenging problems and their potential applications. The 2024 IEEE International Conference on Smart Mobility features 3 tracks representing the smart mobility triad; authors are encouraged to submit papers that fall under the following topic areas: 1) Smart Mobility Technologies 2) City Planning for Smart Mobility 3) Smart Mobility Governance. Submit your paper! We encourage potential authors and delegates to consider becoming a member of the IEEE and IEEE TEMS in order to gain several benefits, including discounted registration fees at SM for the year 2024. Important Dates 1) Paper submission deadline: May, 31 2) Notification of acceptance: July, 1 3) Camera-ready paper submission: August, 2 Paper Preparation, Submission, and IEEE Xplore 1. Paper Preparation: The technical tracks shall accept regular papers of 6 pages maximum (2 additional pages allowed but at an extra charge in the final submission), 10pt font double column in IEEE format. Authors must present novel perspectives within the general scope of the conference. 2. Paper Submission: - The SM uses EDAS for submission. Authors need to: Create an account (if not already registered) with EDAS at https://edas.info/ Submit the paper through the submission link: https://edas.info/N32279 We will be looking forward to receiving your submissions! Kindly promote 2024 IEEE International Conference on Smart Mobility (SM)  Conference to interested individuals and communities in your vicinity. Best regards, and looking forward to seeing you at the 2024 IEEE International Conference on Smart Mobility (SM) in September! SM Technical Committee

9

The #LatinXinAI #Podcast dropped new episodes! Enjoy the conversation about fundamental Reinforcement Learning research with PhD Pablo Samuel Castro, staff research Software Developer in Google Brain (Google DeepMind) in Montreal. Watch the full episode here: https://buff.ly/3ytmGCY 👈🏼 Subscribe to our YouTube channel for updates on upcoming LatinX in AI podcast episodes. #ArtificialIntelligence #MachineLearning #ReinforcementLearning

2

1 Comment

I will be teaching an ICPSR Summer Workshop, "Sensitivity Analysis: Quantifying the Robustness of Inferences to Alternative Factors or Data", Hybrid (virtual, FtF in Ann Arbor, MI, with asynchronous access to materials). https://lnkd.in/etB9AR8h Most social scientists need to consider the unobserved conditions that could change an inference. This workshop is about the statistical tools (R, Stata or the konfound-it app https://lnkd.in/ep_z-DJS ) to do just that, quantifying the strength of evidence to inform policy, practice, or general social science. The workshop is a full 20 instruction hours of instruction including opportunities to adapt to specific models such as mediation, moderation, diff in diff, logistic regression, hazard/cox regression, multilevel models, differential attrition from treatment groups, etc. Go here to register: https://lnkd.in/etB9AR8h Group discounts available as well. https://lnkd.in/eWRv8G7q

10

🏆Impressive work from all T-RIZE Labs team and Professor Kaiwen Zhang ! The integration of AI Federated Learning and blockchain-based carbon credit tokenization is a testament to how decentralized technologies can unlock new levels of privacy, security, and sustainability. 💚By combining these advanced methodologies, T-RIZE GROUP | RWA Tokenization & Onchain Federated Learning is not only driving innovation but also pioneering a path for scalable, eco-friendly solutions that align with global carbon reduction goals. This is the kind of forward-thinking approach that we expect to see from the world’s leading research hubs—an inspiring example for the tech and sustainability sectors alike. Digital Asset Canton Network IEX Chainlink Chainlink Labs

13

3 Comments

Were RNNs all we needed? In a recent paper, our Borealis AI research team explored this question. Borealis AI intern and Mila - Quebec Artificial Intelligence Institute PhD student Leo F., Borealis AI researchers Fred Tung, hossein hajimirsadeghi, and Mohamed Osama Ahmed, and Leo's supervisor Yoshua Bengio showed that LSTMs and GRUs can match the performance of recent sequence models such as the dominant Transformer model. The key insights include dropping hidden state dependencies, which enables parallelization for fast training (e.g., 175x speedup). Sequence models are prevalent in the financial services AI we do at RBC. Transaction sequences, such as payment data, stock market trades, loyalty rewards or other client interactions are common data sources. We have built ATOM, our Large Transaction Model that forms a foundation model for financial services on top of these transaction sequence data. We're excited about the continued push for more performant models in this space, and our paper on RNNs on sequence data help to show where we can go! https://lnkd.in/gqp6vUpr https://lnkd.in/gHdXCPAu #ai #machinelearning #rnn

382

8 Comments

This video highlights my ongoing MASc Thesis at U of T in the department of ECE supervised by Prof. Stephen Brown and Prof. Kevin Truong. Big thank you to Aman B. a PhD candidate at Caltech CNS in the Thomson group, for editing this video, inspiring many initial ideas for the project, and for helping through several breakthroughs this past week! Aman is a fellow EngSci grad from U of T! Also thank you to the Society for AGI for some very helpful discussions! A big inspiration for this project was the paper “Growing Neural Cellular Automata” (https://lnkd.in/gVcfejvn). The authors put a twist on Von Neumann cellular automata by demonstrating the interaction rules could be learned using neural networks to self-organize into desired 2D pixel patterns. Stephen Wolfram had already shown very simple cellular automata to be Turing Complete, so it was only a matter of time before their full possibilities started being explored. Moving from 2D to 3D, and fixed grids to free-moving cells required the insights of geometric deep learning, beautifully outlined in this proto-book: https://lnkd.in/gV4Yybk7. Geometric deep learning revolves around the idea of symmetry—which I think of as “the ways that something can be transformed without changing.” For example, a circle remains exactly the same when rotated in place. In a very real sense, that's what makes it a circle. In the more familiar case of bilateral symmetry, it's reflections over an axis that leave the object unchanged. Symmetry is the concept behind revolutions in modern mathematics (eg. group theory), physics (eg. gauge theory), and now deep learning (in architectures exploiting invariance and equivariance). CNNs led to Group CNNs (https://lnkd.in/g7AkqSK2), and eventually to the SE(3) Transformer (https://lnkd.in/gcV7MVzc). SE(3) is the special euclidean group, consisting of rigid translations and proper rotations in 3 dimensions—meaning this group inherently captures the structure of euclidean space. An architecture with SE(3) equivariance is indispensable for moving from fixed grids to freely moving cells. The power of Equivariant transformers can be seen through their use in AlphaFold, which outputs the 3D coordinates of all heavy atoms for a given protein (https://lnkd.in/gMSPc8t4...) AlphaFold recently won the Nobel Prize in Chemistry for its groundbreaking advance in protein structure prediction. The video of the frog developing was taken by Jan Van Ijken (https://janvanijken.com). Other inspirations include: Michael Elowitz, Rob Phillips, Karl Friston, John Conway, Carver Mead, Geoffrey Hinton, Richard Feynman, Eric Winfree, John Hopfield. I briefly foreshadowed a little bit of this project on the MLST Podcast a few months back (second half of that episode), and you can find that interview here (https://lnkd.in/gvXhKMKk)

101

10 Comments