Again and again, the mistakes that the Trump Administration makes in handling the coronavirus crisis seem to break in the worst possible direction. The latest example concerns antibody tests, which are meant to show whether someone has had COVID-19, but, in practice, often do not. There are more than two hundred tests out there now, produced by a wide range of companies and labs, with little control over how they are marketed; only a dozen have even gone through the process of getting what's known as an Emergency Use Authorization, or E.U.A., from the Food and Drug Administration, which is less rigorous than a normal approval. The F.D.A. has told other companies that they can go ahead and peddle their tests based on self-reported measures of accuracy to clinics, doctors' offices, businesses, or state and local governments. Some tests are not just imperfect but shoddy; "terrible" is the word one researcher used in describing certain tests to CNN.

LSTM or Long Short Term Memory Networks is a specific type of Recurrent Neural Network (RNN) that is very effective in dealing with long sequence data and learning long term dependencies. In this work, we perform sentiment analysis on a GOP Debate Twitter dataset. To speed up training and reduce the computational cost and time, six different parameter reduced slim versions of the LSTM model (slim LSTM) are proposed. We evaluate two of these models on the dataset. The performance of these two LSTM models along with the standard LSTM model is compared. The effect of Bidirectional LSTM Layers is also studied. The work also consists of a study to choose the best architecture, apart from establishing the best set of hyper parameters for different LSTM Models.

In recent years, academics and investigative journalists have criticized certain commercial risk assessments for their black-box nature and failure to satisfy competing notions of fairness. Since then, the field of interpretable machine learning has created simple yet effective algorithms, while the field of fair machine learning has proposed various mathematical definitions of fairness. However, studies from these fields are largely independent, despite the fact that many applications of machine learning to social issues require both fairness and interpretability. We explore the intersection by revisiting the recidivism prediction problem using state-of-the-art tools from interpretable machine learning, and assessing the models for performance, interpretability, and fairness. Unlike previous works, we compare against two existing risk assessments (COMPAS and the Arnold Public Safety Assessment) and train models that output probabilities rather than binary predictions. We present multiple models that beat these risk assessments in performance, and provide a fairness analysis of these models. Our results imply that machine learning models should be trained separately for separate locations, and updated over time.

Benjamin Lee, a second-year Ph.D. student in the Allen School's Artificial Intelligence group working with professor Daniel Weld, has been named a 2020 Innovator in Residence by the Library of Congress. Now in its second year, the Innovator in Residence program aims to enlist artists, researchers, journalists, and others in developing new and creative ways of using the library's digital collections. During his residency, Lee will apply deep learning to enable the automatic extraction and tagging of photographs and illustrations contained in the more than 15 million newspaper scans comprising the library's Chronicling America collection. His goal is to produce interactive visualizations, searchable by topic, that will make the content more accessible to users and support cultural heritage research. "A primary motivation behind my project is to excite the American public by demonstrating the possibilities of applying machine learning to library collections," Lee explained in an interview posted on the library's blog.

"Deepfakes" are digitally altered images that make incidents appear real when they are not. Such altered files could have broad implications for politics. "Deepfakes" are digitally altered images that make incidents appear real when they are not. Such altered files could have broad implications for politics. Sophisticated fake media hasn't emerged as a factor in the disinformation wars in the ways once feared -- and two specialists say it may have missed its moment.

Horrifying Glimpse Into How DARPA Will "Save" You From COVID-19 and Venezuela Coup Tied Back To Trump (7 May 2020). In my last two posts, The New World Emperor and Wake Up, You're Next, I stated that the main worry in the nCov pandemic is not just the virus - its origins, seriousness, the number of strains, and their forthcoming spread - but how the pandemic will be controlled. I argued that mass vaccines and tracking will involve the transition from computer-based to human-based operating systems. A series of pandemic outbreaks now and in coming years will be followed by successive vaccines, which will implant weaponized AI and nanotechnology on a mass scale, in order to establish brain-machine interfaces around the globe, paired with a cryptocurrency as a reward or punishment system. If you accept this technology into your body, the control of the few over the many will be complete, and the Internet of Thoughts will be born. To understand injectable technologies, see (above) The Last American Vagabond's 7 May 2020 interview with independent journalist, Whitney Webb.

Hong Kong (CNN)Australia has its first "loyal wingman." Boeing Australia presented the country's Air Force on Tuesday with a prototype of a jet-powered drone that they hope will one day fly alongside manned warplanes while bringing artificial intelligence to the battlefield. The Loyal Wingman, at 38-foot-long (11.5 meters) and with a range of 2,000 miles (3,218.6 kilometers), will "use artificial intelligence to fly independently, or in support of manned aircraft, while maintaining safe distance between other aircraft," according to Boeing's website on the project. The drones will be able to engage in electronic warfare as well as intelligence, reconnaissance and surveillance missions and swap quickly between those roles, according to Boeing. The aircraft delivered in Sydney on Tuesday is the first of three prototypes Boeing is producing.

MOSCOW: Russia is reportedly aiming to replace human soldiers with what it deems as faster and more accurate robots on the battlefield, with plans to begin testing a newly developed unmanned armor fighting machine. Citing recent statements by deputy director of Russia's Advanced Research Foundation Vitaly Davydov, the US-based Forbes magazine reported that robotics will be the future of warfare due to their increased speed and accuracy in target selection."Living

The Department of Defense is racing to test and adopt artificial intelligence and machine learning solutions to help sift and synthesize massive amounts of data that can be leveraged by their human analysts and commanders in the field. Along the way, it's identifying many of the friction points between man and machine that will govern how decisions are made in modern war. The Machine Assisted Rapid Repository System (MARS) was developed to replace and enhance the foundational military intelligence that underpins most of the department's operations. Like U.S. intelligence agencies, officials at the Pentagon have realized that data -- and the ability to speedily process, analyze and share it among components – was the future. Fulfilling that vision would take a refresh.

Restricted Boltzmann Machine (RBM) is an energy based, undirected graphical model. It is commonly used for unsupervised and supervised machine learning. Typically, RBM is trained using contrastive divergence (CD). However, training with CD is slow and does not estimate exact gradient of log-likelihood cost function. In this work, the model expectation of gradient learning for RBM has been calculated using a quantum annealer (D-Wave 2000Q), which is much faster than Markov chain Monte Carlo (MCMC) used in CD. Training and classification results are compared with CD. The classification accuracy results indicate similar performance of both methods. Image reconstruction as well as log-likelihood calculations are used to compare the performance of quantum and classical algorithms for RBM training. It is shown that the samples obtained from quantum annealer can be used to train a RBM on a 64-bit `bars and stripes' data set with classification performance similar to a RBM trained with CD. Though training based on CD showed improved learning performance, training using a quantum annealer eliminates computationally expensive MCMC steps of CD.