https://techxplore.com/news/2021-04-deeponet-deep-neural-network-based-approximate.html logotype Topics * Week's top * Latest news * Unread news * Subscribe [ ] Science X Account [ ] [ ] [*] Remember me Sign In Click here to sign in with or Forget Password? Not a member? Sign up Learn more * Automotive * Business * Computer Sciences * Consumer & Gadgets * Electronics & Semiconductors * Energy & Green Tech * Engineering * Hardware * Hi Tech & Innovation * Internet * Machine learning & AI * Other * Robotics * Security * Software * Telecom * * share this! * 119 * 57 * Share * Email 1. Home 2. Computer Sciences 1. Home 2. Machine learning & AI * * * --------------------------------------------------------------------- April 12, 2021 feature DeepONet: A deep neural network-based model to approximate linear and nonlinear operators by Ingrid Fadelli , Tech Xplore DeepONet: A deep neural network-based model to approximate linear and nonlinear operators Credit: Lu et al. Artificial neural networks are known to be highly efficient approximators of continuous functions, which are functions with no sudden changes in values (i.e., discontinuities, holes or jumps in graph representations). While many studies have explored the use of neural networks for approximating continuous functions, their ability to approximate nonlinear operators has rarely been investigated so far. Researchers at Brown University recently developed DeepONet, a new neural network-based model that can learn both linear and nonlinear operators. This computational model, presented in a paper published in Nature Machine Intelligence, was inspired by a series of past studies carried out by a research group at Fudan University. George Em Karniadakis, one of the researchers who carried out the study, told TechXplore, "About five years ago, as I was in class teaching variational calculus, I asked myself if a neural network can approximate a functional (we know it approximates a function). I searched around and found nothing, until one day I stumbled on a paper by Chen & Chen published in 1993, where the researchers achieved functional approximation using a single layer of neurons. Eventually, I also read another paper by the same team on operator regression, which we used as a starting point for our study. Since then, Prof T Chen has contacted me by email and thanked me for discovering his forgotten papers." Inspired by the papers by Chen and Chen at Fudan University, Karniadakis decided to explore the possibility of developing a neural network that could approximate both linear and nonlinear operators. He discussed this idea with one of his Ph.D. students, Lu Lu, who started developing DeepONet. In contrast with conventional neural networks, which approximate functions, DeepONet approximates both linear and nonlinear operators. The model comprises two deep neural networks: one network that encodes the discrete input function space (i.e., branch net) and one that encodes the domain of the output functions (i.e., trunk net). Essentially, DeepONet takes functions as inputs, which are infinite dimensional objects, and maps them to other functions in the output space. DeepONet: A deep neural network-based model to approximate linear and nonlinear operators Credit: Lu et al. "With standard neural networks, we approximate functions, which take data points as input and outputs data points," Karniadakis said. "So DeepOnet is a totally new way of looking at neural networks, as its networks can represent all known mathematical operators, but also differential equations in a continuous output space." [INS::INS] Once it learns a given operator, DeepONet can complete operations and make predictions faster than other neural networks. In a series of initial evaluations, Karniadakis and his colleagues found that it could make predictions in fractions of a second, even those related to very complex systems. "DeepONet can be extremely useful for autonomous vehicles, as it can make predictions in real time," Karniadakis said. "It could also be used as a building block to simulate digital twins, systems of systems, and even complex social dynamical systems. In other words, the networks we developed can represent black-box complex systems after intense offline training." As part of their study, the researchers investigated different formulations of DeepONet's input function space and assessed the impact of these formulations on the generalization error for 16 distinct applications. Their findings are highly promising, as their model could implicitly acquire a variety of linear and nonlinear operators. In the future, DeepONet could have a wide range of possible applications. For instance, it could enable the development of robots that can solve calculus problems or solve differential equations, as well as more responsive and sophisticated autonomous vehicles. "I am now collaborating with labs from the Department of Energy and also with many industries to apply DeepONet to complex applications, e.g., in hypersonics, in climate modeling such as applications to model ice-melting in Antarctica, and in many design applications." --------------------------------------------------------------------- Explore further Research shows the intrinsically nonlinear nature of receptive fields in vision --------------------------------------------------------------------- More information: Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators. Nature Machine Intelligence(2021). DOI: 10.1038/s42256-021-00302-5. Journal information: Nature Machine Intelligence (c) 2021 Science X Network Citation: DeepONet: A deep neural network-based model to approximate linear and nonlinear operators (2021, April 12) retrieved 14 April 2021 from https://techxplore.com/news/ 2021-04-deeponet-deep-neural-network-based-approximate.html This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only. 179 shares * Facebook * Twitter * Email Feedback to editors * Featured * Last Comments * Popular FBI launches an effort to mitigate attacker use of Microsoft Exchange vulnerabilities 6 hours ago 1 Membrane design prevents crossover in polysulfide redox flow batteries 8 hours ago 0 Researchers identify a strategy to achieve large transport gap modulation in graphene 9 hours ago 0 DuckDuckGo can now block the Google Chrome tracking method, FLoC Apr 13, 2021 2 A robot that teaches itself to walk using reinforcement learning Apr 13, 2021 0 --------------------------------------------------------------------- [gif] EU to unveil AI rules to fight Big Brother fears 4 hours ago [gif] FBI launches an effort to mitigate attacker use of Microsoft Exchange vulnerabilities 6 hours ago [gif] Membrane design prevents crossover in polysulfide redox flow batteries 8 hours ago [gif] Researchers identify a strategy to achieve large transport gap modulation in graphene 9 hours ago [gif] Biorobotics lab builds submersible robot snake 11 hours ago [gif] Microsoft defends against new threat to Exchange 15 hours ago [gif] DuckDuckGo can now block the Google Chrome tracking method, FLoC Apr 13, 2021 * Related Stories [gif] Research shows the intrinsically nonlinear nature of receptive fields in vision Nov 24, 2020 [gif] Researchers demonstrate all-optical neural network for deep learning Aug 29, 2019 [gif] Concept whitening: A strategy to improve the interpretability of image recognition models Jan 13, 2021 [gif] Evolvable neural units that can mimic the brain's synaptic plasticity Jan 14, 2021 [gif] Artificial neuron device could shrink energy use and size of neural network hardware Mar 18, 2021 [gif] Artificial intelligence predicts nonlinear ultrafast dynamics in optics Feb 19, 2021 * Recommended for you [gif] EU to unveil AI rules to fight Big Brother fears 4 hours ago [gif] eSpine: A technique to increase the usable lifetime of neuromorphic systems Apr 13, 2021 [gif] People may trust computers more than humans Apr 13, 2021 [gif] A robot that teaches itself to walk using reinforcement learning Apr 13, 2021 [gif] Mapping the mob: GIS technology reveals shape of America's mafia networks Apr 13, 2021 [gif] Artificial nervous system uses light sensing to catch objects like humans do Apr 12, 2021 User comments What do you think about this particular story? Your feedback will go directly to Tech Xplore editors. [ ] [ ] [ ] [ ] [ ] Your message to the editors [ ] Your email (only if you want to be contacted back) [ ] Send Feedback Thank you for taking your time to send in your valued opinion to Science X editors. You can be assured our editors closely monitor every feedback sent and will take appropriate actions. Your opinions are important to us. We do not guarantee individual replies due to extremely high volume of correspondence. E-mail the story DeepONet: A deep neural network-based model to approximate linear and nonlinear operators Your friend's email [ ] Your email [ ] [ ] I would like to subscribe to Science X Newsletter. Learn more Your name [ ] Note Your email address is used only to let the recipient know who sent the email. Neither your address nor the recipient's address will be used for any other purpose. The information you enter will appear in your e-mail message and is not retained by Tech Xplore in any form. [ ] [ ] [ ] [ ] [ ] [ ] [ ] Your message [ ] Send Phys.org logotype Phys.org Daily science news on research developments and the latest scientific innovations MedicalXpress logotype Medical Xpress Medical research advances and health news ScienceX logotype Science X The most comprehensive sci-tech news coverage on the web Newsletters [ ] Subscribe Science X Daily and the Weekly Email Newsletter are free features that allow you to receive your favorite sci-tech news updates in your email inbox Follow us * * * * * Top * Home * Search * Mobile version * Help * FAQ * About * Contact * Science X Account * Sponsored Account * Newsletter * Archive * Android app * iOS app * RSS feeds * Push notification (c) Tech Xplore 2014 - 2021 powered by Science X Network Privacy policy Terms of use Your Privacy This site uses cookies to assist with navigation, analyse your use of our services, and provide content from third parties. By using our site, you acknowledge that you have read and understand our Privacy Policy and Terms of Use. Ok More Information E-mail newsletter [ ] Subscribe Follow us * * * * It appears that you are currently using Ad Blocking software. What are the consequences? x Quantcast