The question seems to come in different flavors: symbolic AI or statistical AI, white box AI or black box AI, model driven or data driven AI, generative or discriminative AI? A recent blog by Rich Sutton adds to the list compute-driven AI versus human-knowledge based AI. A number of ...
The question seems to come in different flavors: symbolic AI or statistical AI, white box AI or black box AI, model driven or data driven AI, generative or discriminative AI? A recent blog ...
RT @hardmaru: Probabilistic Model-Based Reinforcement Learning Using The Differentiable Neural Computer They replaced the MDN-LSTM in World Models with a Neural Turing Machine and tried to use MDN-NTMs (can learn algorithms from data, in theory) for model-based RL. https://t.co/FgBbfroH4S https://t.co/Nc4qdoJjKNOpen
My experiments found that a model learned in a Differentiable Neural Computer outperformed a vanilla LSTM based model, on two gaming environments.
RT @DBahdanau: Are you curious about systematic generalization? Do you like small, carefully controlled studies with intriguing conclusions? Check out our latest paper: https://t.co/xA4tYvdjF9. Code & data at https://t.co/ydCswLGkyW. Work done by @MILAMontreal with the help of @Element_ai https://t.co/BZNDgnq0epOpen
The output of the last FiLM block hNq x undergoes an extra 1 × 1 convolution and max-pooling to produce hq x. MAC network of Hudson & Manning (2018) produces hq x by repeatedly applying a ...
RT @tdietterich: Nice article on AI risks from an international strategic perspective by the always insightful @paul_scharre I strongly support his call for governments/militaries to initiate AI safety conversations. https://t.co/qZH9QVON7ZOpen
The real danger of an AI arms race isn't that another country would win; it's that unsafe technologies would make everyone lose.
@f2harrell @danilobzdok @TheWagRocks @DataSciFact Your post is a quite good categorization of statistical properties of ML methods, but it is *not* a way to separate SL from ML. That distinction, as many other statisticians have come to accept, does not exist: https://t.co/sRbf4KImCsOpen
This is the personal website of a data scientist and machine learning enthusiast with a big passion for Python and open source. Born and raised in Germany, now living in East Lansing, ...
@dan_s_becker Dota is a convenient testbed for pushing the limits of general-purpose deep RL technology. Here's a physical robotics problem we solved using the learning system we wrote for Dota: https://t.co/PQGhabYIR1Open
We've trained a human-like robot hand to manipulate physical objects with unprecedented dexterity.
RT @UMDscience: We're proud to announce the launch of the @UofMaryland Center for Machine Learning. @CapitalOne is an inaugural partner of the center, which will be located in the new Brendan Iribe Center. https://t.co/kInrRSrPW1 #UMDdiscovers #UMDinnovates #MachineLearning https://t.co/4bpoTScDi5Open
Initial funding for the center comes from the College of Computer, Mathematical, and Natural Sciences (CMNS) and UMIACS, which will provide technical and administrative support. An ...
Excited to be presenting a keynote at @EmbVisionSummit, May 20-23, Santa Clara, CA: Why the Future of Machine Learning is Tiny. I’ve attended for many years and always enjoy the practical focus! https://t.co/98R3XMUTu0 #computervision #tinyml #tensorflow #evs19Open
The Future of Computer Vision and Machine Learning is Tiny" at the 2019 Embedded Vision Summit.
RT @dbeyer123: For the curious, here's an interview I did with Michael Osborne on AI, automation and its impact on labor. https://t.co/90f1AaqbqNOpen
By David Beyer
RT @ENERGY: ANNOUNCED TODAY: $20M in funding for artificial intelligence and machine learning R&D projects to: ✅Boost grid resilience, operation, and management. ✅Solve scientific problems and improve #AI predictions. ⬇⬇https://t.co/j6dtP4vKGdOpen
Department of Energy Announces $20 Million for Artificial Intelligence Research
RT @SFResearch: And another PhD intern writes a publication accepted to @iclr2019! @AkhileshGotmare goes into detail on: 1. Cosine learning rate decay 2. Learning rate warmup 3. Knowledge distillation Learn more in this Q&A, and congrats on starting full-time this week! https://t.co/S2pAZIrbEH https://t.co/DhCLzsR1CcOpen
Q&A with Salesforce Research Intern Akhilesh Gotmare on how "Optimization and Machine Learning" led him to ICLR.
In the research paper, “A Closer Look at Deep Learning Heuristics: Learning rate restarts, Warmup and Distillation” Futureforce PhD Intern Akhilesh Gotmare, worked with Research Scientist ...
RT @Khipu_AI: #khipu2019 listed as one of the top AI conferences to attend in 2019! https://t.co/4cC1ocw6OR Applications are open! More details in https://t.co/wK0ZLhCYnCOpen
These are the top AI, machine learning, deep learning, conferences to attend this 2019, which will take place in the US and the world…
I finished the book, here's my take: https://t.co/8UoJV6VvVo via @danieldennett "intelligent designer". #ai #deplearning #EasterSunday https://t.co/Gwmd7VOi72Open
What follows is a review of Daniel C Dennett's book on the Evolution of Minds, accompanied by a few of my thoughts. My son gave me this book in December 2017 and I'm afraid, due to my ...
RT @tommy_da_cat: Fantastic work on Differential learning with gaussian process flows from the group of @samikaski #aistats #MachineLearning https://t.co/WkR6E31eFl https://t.co/oZe7xbLpx7Open
We propose a novel deep learning paradigm of differential flows that learn a stochastic differential equation transformations of inputs prior to a standard classification or regression ...
RT @teropa: "Exploring the Creative Potential of Deep Learning through the Magenta Project" by @ada_rob A great introduction to generative AI models and using them for creative purposes, using examples of work by Magenta and others. Some of my stuff in there too 😊 https://t.co/y6CuC0NuYHOpen
https://ai.google/research/people/104881 Exploring the Creative Potential of Deep Learning through the Magenta Project Over the past 3 years, the increase in...
In this clip, @elonmusk tells @lexfridman that adversarial examples are trivially easily fixed. @karpathy is that your experience at @tesla? @catherineols is that what the neurips adversarial challenge found? https://t.co/4OMIKcP67wOpen
Elon Musk is the CEO of Tesla, SpaceX, Neuralink, and a co-founder of several other companies. This conversation is part of the Artificial Intelligence podca...
RT @udacity: Are you looking for a career in #AI? Here are 13 industries soon to be revolutionized by #AI via @Forbes https://t.co/73gl7w6SMROpen
Experts from Forbes Technology Council offer their predictions on which industries are about to be revolutionized by artificial intelligence (AI).
Excited to announce expansion of our partnership with @Amazon Web Services (AWS) on our latest Machine Learning Nanodegree programs - Intro to ML and ML Engineer! Congratulations to the whole @udacity and AWS team! #Udacity #MachineLearning #DreamBig https://t.co/vQFDpzgOX2Open
Build a solid foundation in Supervised, Unsupervised, Reinforcement, and Deep Learning. Then, use these skills to test and deploy machine learning models in a production environment.
RT @scottniekum: Our new IRL algorithm, T-REX, can significantly outperform a suboptimal demonstrator and does not require a model or inference-time rollouts. We achieve SOTA imitation learning performance (vs. GAIL, BCO, etc.) on a variety of MuJoCo and Atari domains: https://t.co/rEcwTgDrNUOpen
A critical flaw of existing inverse reinforcement learning (IRL) methods is their inability to significantly outperform the demonstrator. This is a consequence of the general reliance of ...