It has been widely recognized that adversarial examples can be easily crafted to fool deep networks, which mainly root from the locally non-linear behavior nearby input examples. Applying mixup in training provides an effective mechanism to improve generalization performance and model robustness against adversarial perturbations, which introduces the globally linear behavior in-between training examples. However, in previous work, the mixup-trained models only passively defend adversarial attacks in inference by directly classifying the inputs, where the induced global linearity is not well exploited. Namely, since the locality of the adversarial perturbations, it would be more efficient to actively break the locality via the globality of the model predictions. Inspired by simple geometric intuition, we develop an inference principle, named mixup inference (MI), for mixup-trained models. MI mixups the input with other random clean samples, which can shrink and transfer the equivalent perturbation if the input is adversarial. Our experiments on CIFAR-10 and CIFAR-100 demonstrate that MI can further improve the adversarial robustness for the models trained by mixup and its variants.

Artificial intelligence (AI) is everywhere: personal digital assistants answer our questions, robo-advisors trade stocks for us, and driverless cars will someday take us where we want to go. AI has penetrated our lives, and its use is exploding in biomedical research and health care--including across all dimensions of cancer research, where the potential applications for AI are vast. Artificial Intelligence (AI) is a computer performing tasks commonly associated with human intelligence. Humans are coding or programing a computer to act, reason, and learn. An algorithm or model is the code that tells the computer how to act, reason, and learn.

The federal government is proposing to establish an agency that will focus on robotics and Artificial Intelligence, the Minister of Science and Technology, Dr Ogbonnaya Onu has said. He disclosed this on Monday when he received a delegation of the Academy of Science in his office in Abuja. Dr Onu noted that when established, the agency will help to improve the quality of research in the nation's universities and industrial laboratories. He said the Academy of Science is well-placed to advise the federal government on issues bordering on Science, Technology and Innovation. The minister also said it was high time Nigeria takes its place among the leading nations of the world in terms of Science, Technology and Innovation.

Artificial intelligence programs are multiplying like rabbits across the federal government. The Defense Department has tested AI for predictive maintenance on vehicles and aircraft. Civilian agencies have experimented with robotic process automation. RPA pilots at the General Services Administration and the IRS helped employees save time on repetitive, low-skill tasks. In February, President Donald Trump signed an executive order expanding his administration's efforts to foster the research and development of artificial intelligence tools in government.

The invasive and expensive diagnosis process of bladder cancer, which is one of the most common and aggressive cancers in the United States, may be soon helped by a novel non-invasive diagnostic method thanks to advances in machine learning research at the San Diego Supercomputer Center (SDSC), Moores Cancer Center, and CureMatch Incorporated. Research scientists Igor Tsigelny and Valentina Kouznetsova have been working on the development of a machine-learning (ML) model that looks at a patient's metabolites and their chemical descriptors. The model accurately classifies the stages of bladder cancer in a patient, according to the researchers. Tsigelny is the lead author on a recently published study in the Metabolomics journal called'Recognition of Early and Late Stages of Bladder Cancer using Metabolites and Machine Learning'. When a patient experiences early symptoms of bladder cancer (e.g., blood in urine, pain during urination, etc.), the current method of diagnosis is often a painful, invasive series of tests.

New policies need to be created for artificial intelligence (AI) in order to govern its use while allowing for innovation, according to the US Chamber's Technology Engagement Center and Center for Global Regulatory Cooperation. "The advent of artificial intelligence will revolutionize businesses of all sizes and industries and has the potential to bring significant opportunities and challenges to the way Americans live and work, said Tim Day, senior vice president, Chamber Technology Engagement Center, in a press release. The principles, "serve as a comprehensive guide to address the policy issues pertaining to AI for federal, state, and local policymakers." The chamber also endorsed the Organization for Economic Co-operation and Development's recommendations for AI. "As leaders in the development and use of AI, the U.S. business community has a strong interest in supporting a global AI ecosystem," said Sean Heather, senior vice president of International Regulatory Affairs, US Chamber of Commerce.

New policies need to be created for artificial intelligence (AI) in order to govern its use while allowing for innovation, according to the US Chamber's Technology Engagement Center and Center for Global Regulatory Cooperation. "The advent of artificial intelligence will revolutionize businesses of all sizes and industries and has the potential to bring significant opportunities and challenges to the way Americans live and work, said Tim Day, senior vice president, Chamber Technology Engagement Center, in a press release. The principles, "serve as a comprehensive guide to address the policy issues pertaining to AI for federal, state, and local policymakers." The chamber also endorsed the Organization for Economic Co-operation and Development's recommendations for AI. "As leaders in the development and use of AI, the U.S. business community has a strong interest in supporting a global AI ecosystem," said Sean Heather, senior vice president of International Regulatory Affairs, US Chamber of Commerce.

AI development has morphed into a geopolitical race with China and the United States in a dead heat to be the victor. While the 20th century may be widely thought of as the American century, the 21st will be defined by the leader in this pivotal technology. At least, that's the impression given by most commentators and echoed in Kai-Fu Lee's recent book, AI Superpowers: China, Silicon Valley, and the New World Order. Lee's book fits into the common Cold War narrative, framing the development of this technology as a new space race, even going so far as calling AlphaGo's victory over Ke Jie in 2017 China's "Sputnik moment." His chapters discuss the advantage of the US vs.

Traffic forecasting approaches are critical to developing adaptive strategies for mobility. Traffic patterns have complex spatial and temporal dependencies that make accurate forecasting on large highway networks a challenging task. Recently, diffusion convolutional recurrent neural networks (DCRNNs) have achieved state-of-the-art results in traffic forecasting by capturing the spatiotemporal dynamics of the traffic. Despite the promising results, adopting DCRNN for large highway networks still remains elusive because of computational and memory bottlenecks. We present an approach to apply DCRNN for a large highway network. We use a graph-partitioning approach to decompose a large highway network into smaller networks and train them simultaneously on a cluster with graphics processing units (GPU). For the first time, we forecast the traffic of the entire California highway network with 11,160 traffic sensor locations simultaneously. We show that our approach can be trained within 3 hours of wall-clock time using 64 GPUs to forecast speed with high accuracy. Further improvements in the accuracy are attained by including overlapping sensor locations from nearby partitions and finding high-performing hyperparameter configurations for the DCRNN using DeepHyper, a hyperparameter tuning package. We demonstrate that a single DCRNN model can be used to train and forecast the speed and flow simultaneously and the results preserve fundamental traffic flow dynamics. We expect our approach for modeling a large highway network in short wall-clock time as a potential core capability in advanced highway traffic monitoring systems, where forecasts can be used to adjust traffic management strategies proactively given anticipated future conditions.

We introduce novel communication strategies in synchronous distributed Deep Learning consisting of decentralized gradient reduction orchestration and computational graph-aware grouping of gradient tensors. Networks (DNN) models and data sets (Dai et al., 2019), the need for efficient distributed machine learning strategies on massively parallel systems is more significant than On small to moderate-scale systems, with 10's - 100's of GPU/TPU accelerators, these scaling inefficiencies can be difficult to detect and systematically optimize due to system noise and load variability. The scaling inefficiencies of data-parallel implementations are most readily apparent on large-scale systems such as supercomputers with 1,000's-10,000's of accelerators. Extending data-parallelism to the massive scale of super-computing systems is also motivated by the latter's traditional workload consisting of scientific numerical simulations (Kent & Kotliar, 2018). NVLink interconnect, supporting a (peak) bidirectional bandwidth of 100 GB/s, where each 3 V100 GPUs are grouped in a ring topology with all-to-all connections to a POWER9 CPU.