Group items tagged

Scientific machine learning is a core component of artificial intelligence and a computational technology that can be trained, with scientific data, to augment or automate human skills. Major research advances will be enabled by harnessing DOE investments in massive amounts of scientific data, software for predictive models and algorithms, high-performance computing (HPC) and networking platforms, and the national workforce. The crosscutting nature of machine learning and AI provides strong motivation for formulating a prioritized research agenda. Scientific Machine Learning and Artificial Intelligence (Scientific AI/ML) will have broad use and transformative effects across the research communities supported by DOE. Accordingly, a 2019 Basic Research Needs workshop report identified six Priority Research Directions. The first three PRDs describe foundational research themes that are common to the development of Scientific AI/ML methods and correspond to the need for domain-awareness, interpretability, and robustness. The other three PRDs describe capability research themes and correspond to the three major use cases of massive scientific data analysis (PRD #4), machine learning-enhanced

A grand challenge in computing is the creation of machines that can proactively interpret and learn from data in real time, solve unfamiliar problems using what they have learned, and operate with the energy efficiency of the human brain. While complex machine-learning algorithms and advanced electronic hardware (henceforth referred to as 'hardware') that can support large-scale learning have been realized in recent years and support applications such as speech recognition and computer vision, emerging computing challenges require real-time learning, prediction, and automated decision-making in diverse domains such as autonomous vehicles, military applications, healthcare informatics and business analytics.

The Dreyfus program for Machine Learning in the Chemical Sciences and Engineering, initiated in 2020, provides funding for innovative projects in any area of Machine Learning (ML) consistent with the Foundation's broad objective to advance the chemical sciences and engineering. The Foundation anticipates that these projects will contribute new fundamental chemical insight and innovation in the field.

The goal of the RF Machine Learning Systems (RFMLS) program is to develop the foundations for applying modern data-driven Machine Learning to the RF Spectrum domain as well as to develop practical applications in emerging spectrum problems, which demand vastly improved discrimination performance over today's hand-engineered RF systems. Ultimately, these innovations will result in a new generation of RF systems that are goal-driven and can learn from data rather than being hand-engineered by experts.

The research opportunity described in this announcement falls under the following sections of the BAA: Appendix 1 "Program Description," * Section I entitled "Expeditionary Maneuver Warfare & Combating Terrorism (Code 30); specific thrusts and focused research area: * Paragraph E. "ONR 30 Decision Support, AI, Machine Learning and Graph Analysis Program," Technology Investment area 3, entitled "Modeling/ Machine Learning/ /Artificial Intelligence" *

The primary objectives of the Reversible/Quantum Machine Learning and Simulation (RQMLS) AIE opportunity are (1) to explore the fundamental limits of reversible quantum annealers; (2) to quantitatively predict the computational utility of these systems for machine learning, simulation, and other important tasks; and (3) to design experimental tests for these predictions that can be carried out on newly fabricated small-scale reversible quantum annealers.

Solid research is the underlying driving force to crystallize fearless creativity and innovation. While we are committed to run in-house research and engineering, we are also excited to collaborate with academic partners to facilitate exploration of new and promising research. The Sony Focused Research Award provides an opportunity for university faculty and Sony to conduct this type of collaborative, focused research. The award provides up to $150K USD* in funds, and may be renewed for subsequent year(s). A list of candidate research topics appears below: - Manipulation Secure Image Sensing - Self-Supervised Learning for Spiking Neural Networks with Event-Based Vision Sensor - Deep Learning & Deep Fusion Towards Automotive Scene Perception - Designing and Implementing Camera ISP Algorithms Using Deep Learning and Computer Vision - Robust Mesh Tracking for Volumetric Capture - Advanced Image Processing Enabled by AI - Novel Actuator - Machine Learning/AI for Wireless Communications - Reconfigurable Reflector Type Materials - Individual Treatment Effect Estimation - Acoustic Metamaterials - Novel Technologies for GaN-based VSCELs - Intelligent Sensing of Patient-Reported Outcomes

Global research and development at Sony enables us to foster innovative ideas, which could ultimately lead to future technology advancements and company growth. In order to speed up and expand the creation of new ideas, we would like to partner with universities. This partnership will help cultivate advanced concepts and fertilize our own research and development. The Sony Faculty Innovation Award provides up to $100K in funds to conduct pioneering research in the areas of visualization; computer vision; machine learning; robotics; communications and networking; RF sensing; audio; speech and natural language processing; human computer interaction; mobility; system software; and LSI and hardware.

The DOE SC program in Advanced Scientific Computing Research (ASCR) hereby announces its interest in making research data and artificial intelligence (AI) models findable, accessible, interoperable, and reusable (FAIR1) to facilitate the development of new AI applications in SC's congressionally authorized mission space, which includes the advancement of AI research and development. In particular, ASCR is interested in supporting FAIR benchmark data for AI; and FAIR frameworks for relating data and AI models. For this FOA, AI is inclusive of, for example, machine learning (ML), deep learning (DL), neural networks (NN), computer vision, and natural language processing (NLP). Data, in this context, are the digital artifacts used to generate AI models and/or employed in combination with AI models during inference. An AI model is an inference method that can be used to perform a "task," such as prediction, diagnosis, or classification. The model is developed using training data or other knowledge. An AI task is the inference activity performed by an artificially intelligent system.

The National Science Foundation (NSF) through its Divisions of Electrical, Communications and Cyber Systems (ECCS); Chemical, Bioengineering, Environmental and Transport Systems (CBET); Civil, Mechanical and Manufacturing Innovation (CMMI); Information and Intelligent Systems (IIS); and Mathematical Sciences (DMS) announces a solicitation on Multimodal Sensor Systems for Precision Health enabled by Data Harnessing, Artificial Intelligence (AI), and Learning. Next-generation multimodal sensor systems for precision health integrated with AI, machine learning (ML), and mathematical and statistical (MS) methods for learning can be envisioned for harnessing a large volume of diverse data in real time with high accuracy, sensitivity and selectivity, and for building predictive models to enable more precise diagnosis and individualized treatments. It is expected that these multimodal sensor systems will have the potential to identify with high confidence combinations of biomarkers, including kinematic and kinetic indicators associated with specific disease and disability. This focused solicitation seeks high-risk/high-return interdisciplinary research on novel concepts, innovative methodologies, theory, algorithms, and enabling technologies that will address the fundamental scientific issues and technological challenges associated with precision health.

This mechanism supports basic research to increase knowledge/understanding through discovery and hypothesis generation, and should focus on providing basic fundamental knowledge that will inform and enable the future development of novel autonomous and/or robotic medical systems to care for wounded soldiers/patients through breakthrough, exploratory research. The objective is focused on addressing the following 1. Autonomous and Unmanned Medical Capability - Identify novel ideas, approaches and research towards the conceptualization of autonomous and unmanned technologies for next-generation, high-quality medical capabilities with limited or absent medical care personnel, or personnel with limited skills. Research novel concepts, plausible approaches and advanced concept designs using biologically inspired cognitive computing models, machine learning, artificial intelligence, soft robotic semi-autonomous/autonomous resuscitation concepts and advanced applications of information sciences among other innovative, exploratory research towards advancing the state-of-the-art in delivery of forward resuscitative care at the point of injury. 2. Medical Robotics Research - Identify novel ideas, approaches and research towards the conceptualization of medical robotics and real-time tele-presence capabilities exploring the limits of machine perception for tele-robotic semi-autonomous and autonomous trauma care within remote and dispersed geographic settings. This could include exploratory research in semi-autonomous robotic surgery to improve the safety profile and efficacy of tele-surgical procedures and outcomes using hard robotics in challenging situations (e.g., combat casualties on the multi-domain battlefield or mass casualty situations) and remote or austere geographic locations, among other innovative, exploratory research aims and novel concepts.

The need to process large data sets arising in many practical problems require real-time learning from data streams makes high-performance hardware necessary, and yet the very nature of these problems, along with currently known algorithms for addressing them, imposes significant hardware challenges. Current versions of deep-learning algorithms operate by using millions of parameters whose optimal values need to be determined for good performance in real time on high-performance hardware. Conversely, the availability of fast hardware implementations can enable fuller use of Bayesian techniques, attractive for their ability to quantify prediction uncertainty and thus give estimates of reliability and prediction breakdown. The abilities of ML systems to self-assess for reliability and predict their own breakdowns (and also recover without significant ill effects) constitute critical areas for algorithm development as autonomous systems become widely deployed in both decision support and embodied AI agents. Only with attention to these challenges can we construct systems that are robust when they encounter novel situations or degradation and failure of sensors.

The Energy, Power, Control, and Networks (EPCN) Program supports innovative research in modeling, optimization, learning, adaptation, and control of networked multi-agent systems, higher-level decision making, and dynamic resource allocation, as well as risk management in the presence of uncertainty, sub-system failures, and stochastic disturbances. EPCN also invests in novel machine learning algorithms and analysis, adaptive dynamic programming, brain-like networked architectures performing real-time learning, and neuromorphic engineering. EPCN's goal is to encourage research on emerging technologies and applications including energy, transportation, robotics, and biomedical devices & systems. EPCN also emphasizes electric power systems, including generation, transmission, storage, and integration of renewable energy sources into the grid; power electronics and drives; battery management systems; hybrid and electric vehicles; and understanding of the interplay of power systems with associated regulatory & economic structures and with consumer behavior.

The Energy, Power, Control, andNetworks (EPCN) Program supports innovative research in modeling, optimization, learning, adaptation, and control of networked multi-agent systems, higher-level decision making, and dynamic resource allocation, as well as risk management in the presence of uncertainty, sub-system failures, and stochastic disturbances. EPCN also invests in novel machine learning algorithms and analysis, adaptive dynamic programming, brain-like networked architectures performing real-time learning, and neuromorphic engineering. EPCN’s goal is to encourage research on emerging technologies and applications including energy, transportation, robotics, and biomedical devices & systems. EPCN also emphasizes electric power systems, including generation, transmission, storage, and integration of renewable energy sources into the grid; power electronics and drives; battery management systems; hybrid and electric vehicles; and understanding of the interplay of power systems with associated regulatory & economic structures and with consumer behavior.

The Spectrum Collaboration Challenge will develop intelligent radio networks which can collaborate to manage and optimize the radio frequency (RF) spectrum in a complex dynamic, changing RF environment which consists of other collaborative radio networks, non-collaborative radio networks (which are incapable of adapting) and other potential interference sources. Successful networks will apply machine learning techniques and be able to optimize total spectrum usage by determining when, where and how to utilize its resources. The networks for all participant teams will be evaluated in a series of competitive, tournament-style events.

The Spectrum Collaboration Challenge will develop intelligent radio networks which can collaborate to manage and optimize the radio frequency (RF) spectrum in a complex dynamic, changing RF environment which consists of other collaborative radio networks, non-collaborative radio networks (which are incapable of adapting) and other potential interference sources. Successful networks will apply machine learning techniques and be able to optimize total spectrum usage by determining when, where and how to utilize its resources. The networks for all participant teams will be evaluated in a series of competitive, tournament-style events.

NSF has long supported transformative research in artificial intelligence (AI) and machine learning (ML). The resulting innovations offer new levels of economic opportunity and growth, safety and security, and health and wellness. At the same time, broad acceptance of large-scale deployments of AI systems relies critically on their trustworthiness which, in turn, depends upon the collective ability to ensure, assess, and ultimately demonstrate the fairness, transparency, explainability, and accountability of such systems. Importantly, the beneficial effects of AI systems should be broadly available across all segments of society. NSF and Amazon are partnering to jointly support computational research focused on fairness in AI, with the goal of contributing to trustworthy AI systems that are readily accepted and deployed to tackle grand challenges facing society. Specific topics of interest include, but are not limited to transparency, explainability, accountability, potential adverse biases and effects, mitigation strategies, validation of fairness, and considerations of inclusivity. Funded projects will enable broadened acceptance of AI systems, helping the U.S. further capitalize on the potential of AI technologies. Although Amazon provides partial funding for this program, it will not play a role in the selection of proposals for award.

The Air Force Office of Scientific Research (AFOSR) seeks unclassified proposals from educational institutions in the United States for a University Center of Excellence (UCoE) in Efficient and Robust Machine Learning (ERML). Proposals must not contain any proprietary information. This center is a joint project between AFOSR and the Air Force Research Laboratory, Information Directorate (AFRL/RI; http://www.wpafb.af.mil/afrl/ri.aspx). The center will extend the research capabilities of the Air Force Research Laboratory, and provide opportunities for a new generation of United States scientists and engineers to address the basic research needs of the Air Force. We will consider proposals for up to five (5) years with a three-year (3) base period and a two-year (2) option period. The total anticipated amount for the award is $4M. Each of the three (3) years base is anticipated to be funded at $1M each and $500K each for option years. All funding projections are based on availability of funds. We will evaluate proposals using peer review panels and the criteria specified in section F. Application Review Information. While AFOSR reserves the right to selected and fund all, some, or none of the proposals, we anticipate making one Corporative Agreement award under this competition.

AFOSR seeks unclassified proposals from educational institutions in the United States for a University Center of Excellence (UCoE) in Efficient and Robust Machine Learning (ERML). Proposals must not contain any proprietary information.

The Defense Advanced Research Projects Agency (DARPA) Defense Sciences Office (DSO) is sponsoring a Proposers Day to provide information to potential proposers on the objectives of an anticipated Broad Agency Announcement (BAA) for the Competency-Aware Machine Learning (CAML) program. The Proposers Day will be held via prerecorded webcast on February 20, 2019 at 11:00AM and will repost at 3:00 PM. Advance registration is required for viewing the webcast. Note, all times listed in this announcement and on the registration website are Eastern Time.

The purpose of this program is to develop next generation health care solutions and encourage existing and new research communities to focus on breakthrough ideas in a variety of areas of value to health, such as sensor technology, networking, information and machine learning technology, decision support systems, modeling of behavioral and cognitive processes, as well as system and process modeling. Effective solutions must satisfy a multitude of constraints arising from clinical/medical needs, social interactions, cognitive limitations, barriers to behavioral change, heterogeneity of data, semantic mismatch and limitations of current cyberphysical systems. Such solutions demand multidisciplinary teams ready to address technical, behavioral and clinical issues ranging from fundamental science to clinical practice.

The purpose of this program is to develop next generation health care solutions and encourage existing and new research communities to focus on breakthrough ideas in a variety of areas of value to health, such as sensor technology, networking, information and machine learning technology, decision support systems, modeling of behavioral and cognitive processes, as well as system and process modeling. Effective solutions must satisfy a multitude of constraints arising from clinical/medical needs, social interactions, cognitive limitations, barriers to behavioral change, heterogeneity of data, semantic mismatch and limitations of current cyberphysical systems. Such solutions demand multidisciplinary teams ready to address technical, behavioral and clinical issues ranging from fundamental science to clinical practice.