We present a workflow for clinical data analysis that relies on Bayesian Structure Learning (BSL), an unsupervised learning approach, robust to noise and biases, that allows to incorporate prior medical knowledge into the learning process and that provides explainable results in the form of a graph showing the causal connections among the analyzed features. The workflow consists in a multi-step approach that goes from identifying the main causes of patient's outcome through BSL, to the realization of a tool suitable for clinical practice, based on a Binary Decision Tree (BDT), to recognize patients at high-risk with information available already at hospital admission time. We evaluate our approach on a feature-rich COVID-19 dataset, showing that the proposed framework provides a schematic overview of the multi-factorial processes that jointly contribute to the outcome. We discuss how these computational findings are confirmed by current understanding of the COVID-19 pathogenesis. Further, our approach yields to a highly interpretable tool correctly predicting the outcome of 85% of subjects based exclusively on 3 features: age, a previous history of chronic obstructive pulmonary disease and the PaO2/FiO2 ratio at the time of arrival to the hospital.

Many businesses and industries nowadays rely on large quantities of time series data making time series forecasting an important research area. Global forecasting models that are trained across sets of time series have shown a huge potential in providing accurate forecasts compared with the traditional univariate forecasting models that work on isolated series. However, there are currently no comprehensive time series archives for forecasting that contain datasets of time series from similar sources available for the research community to evaluate the performance of new global forecasting algorithms over a wide variety of datasets. In this paper, we present such a comprehensive time series forecasting archive containing 20 publicly available time series datasets from varied domains, with different characteristics in terms of frequency, series lengths, and inclusion of missing values. We also characterise the datasets, and identify similarities and differences among them, by conducting a feature analysis. Furthermore, we present the performance of a set of standard baseline forecasting methods over all datasets across eight error metrics, for the benefit of researchers using the archive to benchmark their forecasting algorithms.

SCI COMMUN ### Conservation President Joe Biden's administration moved last week to overturn a regulation adopted by his predecessor that eliminated sanctions against companies whose operations accidentally kill migratory birds. Trump administration officials had said the fines should be reserved for intentional deaths—an interpretation that broke with long-standing policy and would have prevented, for example, any penalties against the companies responsible for the 2010 Deepwater Horizon oil spill in the Gulf of Mexico that killed as many as 1 million birds. Wildlife biologists have said the original, broader enforcement was essential to curtail steep declines in populations of 1100 bird species covered by the federal Migratory Bird Treaty Act. Biden's administration would replace Trump's regulation with a new one, which could take months to finalize. Conservation advocates have long proposed a permitting program that would protect companies from legal action for accidental deaths if they adopt practices and technology shown to prevent most bird deaths. ### Energy ![Figure][1] CREDITS: (GRAPHIC) N. CARY/ SCIENCE ; (DATA) GLOBAL ENERGY REVIEW 2021 , IEA Electricity production from renewable sources, led by solar photovoltaics and wind, continued years of steady growth globally during the COVID-19 pandemic, a report says. The production grew 7% in 2020, even as overall demand for electricity dropped and generation from fossil fuels declined. This year, as social restrictions ease and demand climbs, renewable electricity is expected to increase by 8%, the International Energy Agency said last month in its Global Energy Review 2021 . Overall, production from renewables, especially in China, is forecast to provide half the total increase in electricity this year. But power from fossil fuels will grow as well, and analysts say the switch away from coal and other carbon-emitting energy sources is not happening fast enough to reduce the effects of global warming. ### COVID-19 In a move that promises to increase the meager supply of COVID-19 vaccines in poorer countries, the World Health Organization (WHO) last week gave a Chinese-made product the green light. Sinopharm, which makes its vaccine by chemically inactivating the pandemic coronavirus, received an Emergency Use Listing (EUL), a designation WHO gives after reviewing efficacy, safety, and manufacturing practices. It allows the COVID-19 Vaccines Global Access (COVAX) Facility, a global consortium promoting equity in vaccine distribution, to purchase and distribute Sinopharm's vaccine. COVAX and Sinopharm are still negotiating a price, and the company says it can increase production. More than 65 million doses have already been administered in 45 countries that have authorized its use. COVAX has struggled to buy affordable vaccines that have EULs—most are too expensive, or supplies were prepurchased by other countries. To date, COVAX has shipped fewer than 60 million doses. ### Vaccines After a protracted internal debate, the Biden administration last week said it supports a proposed international agreement to waive patents on the intellectual property used to make COVID-19 vaccines. The move was hailed by advocates of increased vaccine access and fairness in their distribution, who contend that it will attract new companies to help alleviate a global shortage and reduce costs. But many health officials and vaccinemakers caution that it will not increase supply for many months and that newcomers who want to produce the vaccines face bottlenecks including a lack of technical know-how and widespread shortage of raw materials. ### Public health U.S. regulators on 10 May authorized Pfizer and BioNTech's COVID-19 vaccine for use in children ages 12 to 15, expanding availability beyond older teens and adults. The go-ahead is the first in the United States for this age group and a key step in restarting in-person schooling, team sports, and other group activities. The decision by the U.S. Food and Drug Administration (FDA) was expected after the companies announced in March results from a trial of 2260 adolescents in that age group; 18 who received a placebo developed COVID-19 versus none who received the vaccine. Pfizer has said it expects to obtain safety and efficacy data from clinical trials studying children ages 2 to 11 by September, when it plans to ask FDA to permit use in that age group. A small fraction of all Americans who have died from COVID-19 were under age 18, and children who contract COVID-19 tend to have milder symptoms, but some face long-term health problems. However, many U.S. parents remain hesitant to vaccinate their children because of misinformation and because FDA has only authorized COVID-19 vaccines for emergency use, not given them full approval. ### Virology The World Health Organization (WHO) on 10 May designated as a variant of concern a version of the pandemic coronavirus first identified in India in February. Evidence suggests the variant, B.1.617, now found in about 40 countries, is more transmissible than original strains of SARS-CoV-2, WHO said. It is the fourth variant to receive this WHO designation, following ones first found in Brazil, South Africa, and the United Kingdom. ### Governance To be better prepared for the next pandemic, the world needs a Global Health Threats Council akin to the United Nations Security Council, which would bring together country leaders from different regions along with representatives of the private sector and civil society, according to the first comprehensive review of the global response to COVID-19, released on 12 May. The 13-member Independent Panel for Pandemic Preparedness and Response, which was commissioned by the World Health Organization, also proposes creating an International Pandemic Financing Facility with annual funding of $5 billion to $10 billion and giving WHO a bigger budget, more independence, and new powers to investigate outbreaks anywhere in the world, among many other recommendations. The panel was led by two former heads of state, Ellen Johnson Sirleaf of Liberia and Helen Clark of New Zealand, and included public health specialists, diplomats, and economists. “Pandemics pose potential existential threats to humanity and must be elevated to the highest level,” says the report, COVID-19: Make it the Last Pandemic . ### Public attitudes U.S. adults voice varying support for rules by governments and businesses that could require people to prove they received a COVID-19 vaccine before certain in-person activities, such as attending crowded events or entering their office, a Gallup poll reported last week. Most of the 3731 respondents surveyed in April favored proof to travel by airplane (57%) and attend large events such as concerts (55%). But support dipped below a majority for requiring proof to enter one's workplace (45%) or dine indoors at a restaurant (40%). Responses differed by political party affiliation and willingness to be vaccinated. ### Leadership Nancy Messonnier, a top official at the U.S. Centers for Disease Control and Prevention (CDC) who drew the ire of former President Donald Trump in February 2020 for her unvarnished public warning about the impending coronavirus pandemic, will leave the agency on 14 May, she told colleagues last week in an email. Two weeks before the email, she had been reassigned from heading CDC's COVID-19 task force, Politico reported. She will join the Skoll Foundation as executive director for pandemics and health systems. At CDC since 1995, Messonnier led the launch of a low-cost meningitis vaccine in Africa and rose to head the agency's National Center for Immunization and Respiratory Diseases. After she warned at a press conference 15 months ago that the pandemic might severely sicken many Americans, stock markets plunged, and she did not appear at any subsequent White House briefings. ### Seismology One of the largest seismic research projects in history, used to study structures in Earth's crust and mantle as deep as 3000 kilometers, is ending its run. Begun in 2004 and funded by the U.S. National Science Foundation, the several hundred seismic stations of the Transportable Array collected earthquake data as a way of peering deep inside Earth. The stations were deployed at first in the western states and then were moved east across the country every few years. In 2017, the project's final phase began when the network was transported to Alaska. Pandemic-related delays prompted an extra year of operation, but in early May, more than 80 stations went dark, awaiting collection this summer. Another 100 will remain in Alaska, many of them in remote, previously unmonitored regions, filling gaps in the state's seismic coverage. ### Drug development A psychedelic drug has passed a major milestone by showing evidence of benefit as a supplement to talk therapy for post-traumatic stress disorder (PTSD), researchers report this week. The results came from the first phase 3 clinical trial combining psychotherapy with the drug 3,4-methylenedioxymethamphetamine (MDMA), popularly called ecstasy, for people with severe, chronic PTSD. The study, sponsored by the Multidisciplinary Association for Psychedelic Studies (MAPS), recruited 90 people to receive talk therapy during 15 sessions, including three “experimental” ones in which they received either MDMA or a placebo. The MDMA group saw significantly greater improvements on a PTSD symptom scale, the researchers said in Nature Medicine . Two months after the final experimental session, 67% of those who got MDMA no longer met the diagnostic criteria for PTSD, versus 32% of those who got a placebo. MAPS aims to confirm those benefits in a 100-person trial now enrolling volunteers and, in 2023, to seek approval from the U.S. Food and Drug Administration for MDMA-assisted therapy. ### Conservation Researchers are pressing Chile's government to increase protections for whales along its coast against deadly ship collisions. After three whales were found dead during just 8 days in April, 65 Chilean marine mammal specialists issued a public plea for the government to act. They called for rerouting ships away from sensitive regions, setting speed limits, and establishing an alert system to warn vessel pilots of nearby whales. Hundreds of vessels plying Chilean waters pose significant threats to the estimated 40% of the world's cetacean species that frequent them, researchers say. In 2008, Chile declared its entire 6500-kilometer-long coastline a whale sanctuary. But the country's protections for marine mammals still exist only on paper or lack details. The government says it is working to increase safety measures. But researchers say it must commit adequate funding to succeed. ### Space science NASA marked two milestones this week—the final tests on Earth of its next marquee space telescope, and the start of a long journey home for a trove of asteroid rocks. NASA engineers are wrapping up testing on the giant gold-tinted mirror of the James Webb Space Telescope. Afterward, the long-delayed, $9 billion observatory will be shipped off to French Guiana for launch on 31 October. Also this week, NASA's OSIRIS-REx probe began its 2-year flight back to Earth from the asteroid Bennu, carrying up to 400 grams of rocky chunks the spacecraft collected from its surface. It will be the largest U.S. haul of rocks collected in space since the Apollo program, and NASA's first from an asteroid. ### Aquaculture Nutrition expert Shakuntala Haraksingh Thilsted was awarded the World Food Prize this week for research that used aquaculture to improve the diets of millions of people across Asia and Africa. Thilsted, who began her career at Trinidad and Tobago's Ministry of Agriculture, Land & Fisheries, now heads nutrition research at WorldFish, a nonprofit research center in Penang state in Malaysia. In Bangladesh, she studied small, native fish species widely eaten by farmers. Thilsted identified micronutrients the fish contain and their valuable role in the healthy development of infants and toddlers; adding these fish also helps the body absorb other micronutrients in rice and vegetables. She developed ways to cheaply raise fish, combining large and small species in ponds, which increases production. These methods have helped make Bangladesh one of the world's top aquaculture producers. [1]: pending:yes

In the past 10 years, hundreds of projects have applied artificial intelligence (AI) to creating social good. The right tool applied to an appropriate problem has the potential to drastically improve millions of lives through better service delivery and better-informed policy design. But what kind of investments do AI solutions need to be successful, and which applications have the most potential for social impact? AI excels at helping humans harness large-scale or complex data to predict, categorize, or optimize at a scale and speed beyond human ability. We believe that more targeted, sustained investments in AI for social impact (sometimes called "AI for good")--rather than multiple, short-term grants across a variety of areas--are important for two reasons.

Yext uses artificial intelligence (AI) to help companies with their digital transformation in response to the Covid-19 lockdown. Its biggest customers are mobile phone operators such as Three and Verizon. It helps these communications service providers (CSPs) by improving the digital experience of customers. For many mobile phone users, the customer journey often starts with a search for information about one supplier and ends with them going to a rival which provided the answers more readily. "If they want customers to help themselves, they need to structure their information so the answers are readily available," says Jon Buss, managing director of Europe, the Middle East and Africa (EMEA) for AI company Yext.

On the As one of the solutions to the Dec-POMDP problem, the value other hand, in order to extract helpful information from the state of decomposition method has achieved good results recently. However, the complex environment, some work[12, 19] promotes the neural most value decomposition methods require the global state network to learn useful state information by adding auxiliary tasks during training, but this is not feasible in some scenarios where mainly to predict the state of the next moment. Intuitively, the the global state cannot be obtained. Therefore, we propose a novel problem with these studies is in that they cannot be implemented value decomposition framework, named State Inference for value for tasks that cannot obtain real state information. DEcomposition (SIDE), which eliminates the need to know the true As a notorious problem in MAS, Dec-POMDP[25] describes some state by simultaneously seeking solutions to the two problems of collaboration problems.

Each artwork is 512 x 512 pixels and the owner will receive the full resolution at 51200 x 25600 pixels (1.3 Giga Pixel).

Scientists have developed software that can look minutes into the future. Gravity in this galaxy [even outside our solar system] behaves as predicted by Albert Einstein's general theory of relativity, confirming the theory's validity on galactic scales. FYI: Our Sun is just ONE STAR among the hundreds of billions of stars in our Milky Way Galaxy. The technology is being built into the official postal system in countries like Mongolia, Ivory Coast and Nigeria, and Mercedes, an investor, is incorporating "What3Words" navigation into its cars. FYI: Each 10-foot-square patch of Planet Earth is labeled with three words -- 57 trillion squares altogether.

In this paper we show how using satellite images can improve the accuracy of housing price estimation models. Using Los Angeles County's property assessment dataset, by transferring learning from an Inception-v3 model pretrained on ImageNet, we could achieve an improvement of ~10% in R-squared score compared to two baseline models that only use non-image features of the house.

A key question in many low-rank problems throughout optimization, machine learning, and statistics is to characterize the convex hulls of simple low-rank sets and judiciously apply these convex hulls to obtain strong yet computationally tractable convex relaxations. We invoke the matrix perspective function - the matrix analog of the perspective function-and characterize explicitly the convex hull of epigraphs of convex quadratic, matrix exponential, and matrix power functions under low-rank constraints. Further, we exploit these characterizations to develop strong relaxations for a variety of low-rank problems including reduced rank regression, non-negative matrix factorization, and factor analysis. We establish that these relaxations can be modeled via semidefinite and matrix power cone constraints, and thus optimized over tractably. The proposed approach parallels and generalizes the perspective reformulation technique in mixed-integer optimization, and leads to new relaxations for a broad class of problems.