While AI and machine learning have the potential for transforming healthcare, the technology has inherent biases that could negatively impact patient care, senior FDA officials and Philips' head of global software standards said at the meeting. Bakul Patel, director of FDA's new Digital Health Center of Excellence, acknowledged significant challenges to AI/ML adoption including bias and the lack of large, high-quality and well-curated datasets. "There are some constraints because of just location or the amount of information available and the cleanliness of the data might drive inherent bias. We don't want to set up a system and we would not want to figure out after the product is out in the market that it is missing a certain type of population or demographic or other other aspects that we would have accidentally not realized," Patel said. Pat Baird, Philips' head of global software standards, warned without proper context there will be "improper use" of AI/ML-based devices that provide "incorrect conclusions" provided as part of clinical decision support.

Three Albuquerque-based companies and New Mexico State University will compete alongside nine out-of-state entities for $50,000 in cash prizes in this year's Hyperspace Challenge. Organizers of the challenge, now in its third year, selected a total of 11 companies and two universities to participate in the 2020 accelerator program, which will focus on developing new, innovative technology to help the U.S. Space Force provide satellites and spacecraft with remote, autonomous ability to manage problems. The Air Force Research Laboratory at Kirtland Air Force Base launched the annual challenge in 2018 in partnership with the ABQid business accelerator run by CNM Ingenuity. The program pairs participating companies with government contractors to resolve critical issues, potentially leading to contracts to build new technology for the U.S. Department of Defense and other federal entities. The last two accelerators in 2018 and 2019 focused, respectively, on data analytics to manage reams of information received from space operations, and new technologies for small satellites.

Space robotics startup GITAI and the Japan Aerospace Exploration Agency (JAXA) are teaming up to produce the world's first robotics demonstration in space by a private company. The new agreement under the JAXA Space Innovation through Partnership and Co-creation (J-SPARC) initiative aims to demonstrate the potential for robots to automate of the processing of specific tasks aboard the International Space Station (ISS). Robotics is altering many aspects of our lives in many fields and one where it is particularly attractive is in the exploration and exploitation of space. Ironically, the great strides made in manned spaceflight since the first Vostok mission lifted off in 1961 have shown that not only is supporting astronauts in orbit challenging and expensive, there are also many tasks, like microgravity experiments, where the human touch isn't the best choice. These tasks often require complex, precise, and subtle movements that demand either a highly specialized and expensive bespoke apparatus or a robot.

Humans have been teaming up with machines throughout history to achieve goals, be it by using simple machines to move materials or complex machines to travel in space. But advances in artificial intelligence today bring possibilities for even more sophisticated teamwork -- true human-machine teams that cooperate to solve complex problems. Much of the development of these human-machine teams focuses on the machine, tackling the technology challenges of training AI algorithms to perform their role in a mission effectively. But less focus, MIT Lincoln Laboratory researchers say, has been given to the human side of the team. What if the machine works perfectly, but the human is struggling?

Amid NASA's progress in AI research starting from ML model to predict hurricanes to partnering with Google to make quantum computing accessible, it has now developed a new AI tool to classify a cluster of craters on Mars. The launch of this new AI tool, built on a machine learning algorithm, was aimed at helping scientists to reduce their process time of scanning a single Context Camera image. Thus, researchers from Jet Propulsion Laboratory (JPL), created this tool also called an "automated fresh impact crater classifier", where for the "first time" researchers are leveraging AI to identify unknown craters on the Red Planet, stated by NASA, in their statement. According to their news release, typically scientists and researchers spend hours each day studying images to understand "dust devils, avalanches, and shifting dunes," and approximately 40 minutes to scan a single Context Camera image; however this tool will significantly reduce the processing time and advance the workflow massively. The launch of this tool is a part of a broader NASA's bigger effort -- COSMIC -- capturing onboard summarization to monitor image change that develops technologies for future generations of Mars orbiters.

L3Harris Technologies will help the U.S. Department of Defense (DOD) develop artificial intelligence and machine learning (AI/ML) systems to help reduce the amount of time it takes to decipher usable intelligence from increasing amounts of data collected from space and airborne assets. L3Harris will research, develop and demonstrate an AI/ML interface using data science techniques under a new multimillion-dollar contract to support DOD applications. "L3Harris' work will allow the DOD to turn massive volumes of data into actionable intelligence," said Ed Zoiss, president, Space and Airborne Systems, L3Harris. "The abundance of data collected by space and airborne assets is only increasing. The findings of this research will directly address the data processing challenges within the DOD and intelligence community."

Continual labeling of training examples is a costly task in supervised learning. Active learning strategies mitigate this cost by identifying unlabeled data that are considered the most useful for training a predictive model. However, sample selection via active learning may lead to an exploitation-exploration dilemma. In online settings, profitable items can be neglected when uncertain items are annotated instead. To illustrate this dilemma, we study a human-in-the-loop customs selection scenario where an AI-based system supports customs officers by providing a set of imports to be inspected. If the inspected items are fraud, officers levy extra duties, and these items will be used as additional training data for the next iterations. Inspecting highly suspicious items will inevitably lead to additional customs revenue, yet they may not give any extra knowledge to customs officers. On the other hand, inspecting uncertain items will help customs officers to acquire new knowledge, which will be used as supplementary training resources to update their selection systems. Through years of customs selection simulation, we show that some exploration is needed to cope with the domain shift, and our hybrid strategy of selecting fraud and uncertain items will eventually outperform the performance of the exploitation strategy.

Such examples include predicting input/output responses, design, and optimization [55]. These ODEs and their solutions often depend on a set of input parameters, and such ODEs are denoted as parameterized ODEs. Examples of such input parameters within the context of fluid dynamics include Reynolds number and Mach number. In many important scenarios, high-fidelity solutions of parameterized ODEs are required to be computed i) for many different input parameter instances (i.e., many-query scenario) or ii) in real time on a new input parameter instance. A single run of a high-fidelity simulation, however, often requires fine spatiotemporal resolutions. Consequently, performing real-time or multiple runs of a high-fidelity simulation can be computationally prohibitive. To mitigate this computational burden, many model-order reduction approaches have been proposed to replace costly high-fidelity simulations. The common goal of these approaches is to build a reduced-dynamical model with lower complexity than that of the high-fidelity model, and to use the reduced model to compute approximate solutions for any new input parameter instance. In general, model-order reduction approaches consist of two components: i) a low-dimensional latent-dynamics model, where the computational complexity is very low, and ii) a (non)linear mapping that constructs high-dimensional approximate states (i.e., solutions) from the low-dimensional states obtained from the latent-dynamics model.

A novel extension of the Probabilistic Learning on Manifolds (PLoM) is presented. It makes it possible to synthesize solutions to a wide range of nonlinear stochastic boundary value problems described by partial differential equations (PDEs) for which a stochastic computational model (SCM) is available and depends on a vector-valued random control parameter. The cost of a single numerical evaluation of this SCM is assumed to be such that only a limited number of points can be computed for constructing the training dataset (small data). Each point of the training dataset is made up realizations from a vector-valued stochastic process (the stochastic solution) and the associated random control parameter on which it depends. The presented PLoM constrained by PDE allows for generating a large number of learned realizations of the stochastic process and its corresponding random control parameter. These learned realizations are generated so as to minimize the vector-valued random residual of the PDE in the mean-square sense. Appropriate novel methods are developed to solve this challenging problem. Three applications are presented. The first one is a simple uncertain nonlinear dynamical system with a nonstationary stochastic excitation. The second one concerns the 2D nonlinear unsteady Navier-Stokes equations for incompressible flows in which the Reynolds number is the random control parameter. The last one deals with the nonlinear dynamics of a 3D elastic structure with uncertainties. The results obtained make it possible to validate the PLoM constrained by stochastic PDE but also provide further validation of the PLoM without constraint.

Demographic inference from text has received a surge of attention in the field of natural language processing in the last decade. In this paper, we use personal names to infer religion in South Asia - where religion is a salient social division, and yet, disaggregated data on it remains scarce. Existing work predicts religion using dictionary based method, and therefore, can not classify unseen names. We use character based models which learn character patterns and, therefore, can classify unseen names as well with high accuracy. These models are also much faster and can easily be scaled to large data sets. We improve our classifier by combining the name of an individual with that of their parent/spouse and achieve remarkably high accuracy. Finally, we trace the classification decisions of a convolutional neural network model using layer-wise relevance propagation which can explain the predictions of complex non-linear classifiers and circumvent their purported black box nature. We show how character patterns learned by the classifier are rooted in the linguistic origins of names.