We present an approach to data fusion that combines the interpretability of structured probabilistic graphical models with the flexibility of neural networks. The proposed method, lightweight data fusion (LDF), emphasizes posterior analysis over latent variables using two types of information: primary data, which are well-characterized but with limited availability, and auxiliary data, readily available but lacking a well-characterized statistical relationship to the latent quantity of interest. The lack of a forward model for the auxiliary data precludes the use of standard data fusion approaches, while the inability to acquire latent variable observations severely limits direct application of most supervised learning methods. LDF addresses these issues by utilizing neural networks as conjugate mappings of the auxiliary data: nonlinear transformations into sufficient statistics with respect to the latent variables. This facilitates efficient inference by preserving the conjugacy properties of the primary data and leads to compact representations of the latent variable posterior distributions. We demonstrate the LDF methodology on two challenging inference problems: (1) learning electrification rates in Rwanda from satellite imagery, high-level grid infrastructure, and other sources; and (2) inferring county-level homicide rates in the USA by integrating socio-economic data using a mixture model of multiple conjugate mappings.

JOHANNESBURG, South Africa – 19 November, 2020 – Samsung has announced that it will hold the Samsung AI Forum 2020 online via its YouTube channel for two days from November 2nd to 3rd. Marking its fourth anniversary this year, the forum gathers world-renowned academics and industry experts on artificial intelligence (AI) and serves as a platform for exchanging ideas, insights and latest research findings, as well as a platform to discuss the future of AI. On Day 1, which will be hosted by Samsung Advanced Institute of Technology (SAIT), Samsung's R&D hub dedicated to cutting-edge future technologies, Dr. Kinam Kim, Vice Chairman & CEO of Device Solutions at Samsung Electronics will deliver opening remarks. Renowned AI experts will subsequently give presentations under the theme "AI Technologies for Changes in the Real World." This year, Dr. Inyup Kang, President of System LSI Business at Samsung Electronics will join the panel discussion with the presenters.

The current pandemic has seen organisations accelerating their investment in digital technology strategies to cope with new ways of working. It has also been an opportunity for businesses to rethink their manufacturing models and the skills required now and in the future. Artificial Intelligence (AI) in manufacturing is a natural evolution of the current processes commonly found in automation. Many manufacturing lines being used within supply chains are already using algorithms that apply AI to determine optimum running sequences for lines. Additionally, allocating complex customer orders and logistics deliveries could use similar forms of algorithms.

Written communication is among the greatest inventions in human history, yet reading and writing are skills most of us take for granted. After we learn them at school, we seldom stop to think about the mental-cum-physical process that turns our language and thoughts into symbols on a piece of paper or computer screen, or the reverse process whereby our brains extract meaning from written symbols. The neural correlates of reading remain a mystery to neuroscientists. They once assumed that an auditory pathway in the brain was used for alphabetic symbols and a visual pathway for Chinese characters but have since discovered experimentally that both neural pathways are used together—if in differing proportions—in each instance. Meanwhile, key aspects of writing's development have yet to be demystified by archaeologists and philologists. Was there a single origin, circa 3100 BCE—either cuneiform in Mesopotamia or hieroglyphs in Egypt—or did writing arise in multiple places independently? When and how did Chinese characters, first identified on Shang oracle bones dated to circa 1200 BCE, originate? And what prompted the invention of the radically simple alphabetic principle, circa 1800 BCE, in a script that contains certain signs resembling Egyptian hieroglyphs? The Secret History of Writing —a BBC television series broadcast in three parts, two of which have been adapted as NOVA's A to Z: The First Alphabet and A to Z: How Writing Changed the World —explores these questions and more. Both versions of the series are intelligent, articulate, and visually imaginative, discussing five millennia of writing—by hand, by printing, and by computer keyboard. The programs feature notable scholars of many scripts and cultures, such as Assyriologist Irving Finkel, Egyptologist Pierre Tallet, and Sinologist Yongsheng Chen, interviewed by Lydia Wilson, an academic with expertise in medieval Arabic philosophy and the winning ability to interrogate authorities at their own level while rendering their views broadly understandable and engaging. The idea for the series grew from a long-standing friendship between writer-director David Sington and calligrapher Brody Neuenschwander, who charismatically demonstrates his skill at penning ancient and modern scripts, using materials such as Egyptian papyrus, European parchment, and Islamic paper. At one point, Neuenschwander observes that Latin alphabetic letter forms, unlike calligraphic scripts such as Chinese and Arabic, were ideally shaped for the movable metal type created by Johannes Gutenberg in the 1450s—a technology that enabled the growth of European literacy and the European scientific revolution beginning in the 16th century. The pairing was so ideal, in fact, that the Gutenberg Bible fooled some scholars for centuries, who believed it was handwritten and cataloged it as such. “I think Gutenberg would have been delighted by our confusion, because what he was trying to achieve with the printing of this book was to produce a book, by a new technique, that people would think was just as good as the manuscripts that they were used to buying and reading,” observes archivist Giles Mandelbrote. He was trying to do “something new that would seem old.” In another scene, Finkel, a lifelong scholar of cuneiform at the British Museum, avidly dissects a few signs on early clay tablets to explain the rebus principle, which permits the sounds of pictograms, written together, to express the sound of an unrelated, nonpictographic word. Thus, for example, the plainly pictographic Sumerian sign for barley, pronounced “she,” can be written beside the pictographic sign for milk, pronounced “ga,” to create two signs read as “shega,” meaning something like “beautiful.” As Finkel reasonably speculates, rebuses are so “obvious” that they could have been developed in languages anywhere in the world, supporting the hypothesis that writing may have arisen on multiple, separate occasions. Today, pictography has returned to writing in the form of international transport symbols and computerized emojis. Meanwhile, many young people in China, having become habituated to smartphone writing, are increasingly using the Romanized spelling known as Pinyin (“spell sound”) and, as a result, some no longer know how to write Chinese characters. Could smartphones, or the internet more generally, eventually lead to a universal writing system, independent of particular languages, like the one envisioned by polymath Gottfried Leibniz in 1698? It is unlikely, in my view, and, according to Wilson, undesirable. “A world of perfect communication is also a world of cultural uniformity,” she cautions.

Several infectious diseases cause considerable mortality worldwide each year: Tuberculosis causes ∼1.2 million deaths, diarrheal disease causes ∼1.5 million deaths, and lower respiratory infections cause ∼700,000 deaths in children under 5 years old ([ 1 ][1]). Yet the scale and speed of innovation in developing tools for coronavirus disease 2019 (COVID-19) dwarf the development of those for global infectious diseases, which disproportionally affect resource-limited countries. By August 2020, ∼175 therapeutics and vaccines were in clinical trials for COVID-19 ([ 2 ][2]). By contrast, for 41 global infectious diseases or disease groups, only ∼250 therapeutics and vaccines were in clinical trials in August 2019 ([ 3 ][3]). A robust product pipeline and abridged development time frame for COVID-19 has primarily been enabled by three factors: scientific advances, operational efficiencies, and large-scale at-risk financing. A clear, well-financed path from research through product procurement now exists for COVID-19, shortening timelines while increasing output. This could underpin an approach for global infectious diseases. Recent scientific advances have revolutionized platform technologies and expanded the ability to rapidly identify therapeutic and vaccine candidates. High-throughput computational screening of molecular libraries against key pathogens and/or host targets has accelerated the ability to repurpose agents and identify entities against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2, which causes COVID-19) ([ 4 ][4]). Candidate compounds with existing clinical safety data quickly entered clinical trials, leading to the repurposing of dexamethasone and remdesivir to treat hospitalized COVID-19 patients. Monoclonal antibodies (mAbs) can potentially provide near-immediate therapy and/or prophylaxis by bypassing the need for a host-generated immune response ([ 5 ][5]), and at lower costs and higher volumes than previously assumed. Vaccines have benefited from innovations in vector modalities, manufacturing, antigen design, computational biology, protein engineering, and gene synthesis ([ 6 ][6]). Such innovations may provide the technological basis for targeting other global infectious diseases. In response to COVID-19, the public health and regulatory communities are streamlining clinical development. Independently funded, designed, and conducted platform clinical trials, such as Accelerating COVID-19 Therapeutic Interventions and Vaccines (ACTIV), are structured under a single, adaptive “master” protocol to allow for continuous and consistent evaluation of multiple drug candidates, adding products as they become available and removing candidates as they are deemed futile. They also provide access to large, geographically diverse populations, and some have created or expanded operational structures in resource-limited countries ([ 7 ][7]). Timelines have been shortened because of accelerated regulatory reviews, flexible requirements to enter first-in-human trials, newer approaches to modeling population-specific issues, early approval mechanisms, and enhanced regulatory harmonization among countries ([ 8 ][8]). This increased efficiency in clinical trial execution and regulatory processes could be applied to other global infectious diseases. Historically, investment in product development for global infectious diseases has been restricted owing to the lack of financial returns compared to more profitable areas of drug development, such as oncology. However, the threat that pandemic human coronaviruses (HCoVs) pose to the global economy, political stability, and people's lives has stimulated the private sector, public sector, and philanthropic groups to devote considerable financial and human resources to product development. Previous HCoV outbreaks led to initial development activities that were accelerated with COVID-19. Supplementing these efforts, the U.S. government has provided over $10 billion for COVID-19 therapeutics and vaccines. Other governments, including the European Union, United Kingdom, Germany, and Canada, are making substantial financial commitments, as are large funding institutions ([ 2 ][2]). A fundamental principle behind this unprecedented funding is that financing for the entire product development process is made by the time a candidate enters early-stage clinical trials ([ 9 ][9]). This approach has mitigated the range of risks faced by different categories of developers (e.g., academia, nonprofit organizations, public-private partnerships, small biotechnology companies, and large multinational pharmaceutical companies) who may individually have widely varying risk-reward calculations. As a result, developers can simultaneously prepare for late-stage clinical trials, implement scaling up of manufacturing processes, and obtain advanced purchase commitments of large-scale supply—all during first-in-human clinical trials ([ 9 ][9]). Together, providing the full range of financing as early as possible in the product development process, articulating the need for multiple products, and acknowledging implicit failure of some candidates and platforms have overcome product development barriers. The result has been an extraordinary scale of therapeutic and vaccine development in the shortest time possible. A similar product development framework could be created for global infectious diseases. Such a framework could attempt to resolve three long-standing challenges for these diseases: the lack of interest in developing products, resulting in a diminished initial pipeline of candidates; the large pipeline attrition points between preclinical activities and early-stage clinical trials and between early- and late-stage clinical trials ([ 10 ][10]) that occur because of the considerable increases in development costs of these two transition points; and the extended timelines for product development. If these challenges are addressed, a more robust initial pipeline could be created, more candidates could advance to early- and late-stage clinical trials, and more products could be approved in a shorter period. A robust pipeline for global infectious diseases should include repurposed agents, mAbs, new chemical entities, and vaccines. Each of these categories possess strengths and limitations; thus, each may not prove beneficial for every disease. Repurposed agents may have existing preclinical data and clinical safety experience, putting them on the fastest development timelines. mAbs targeting proteins encoded by highly conserved regions of a pathogen's genome—thereby minimizing escape mutations and maximizing strain coverage—can be isolated from patients and modified to enhance their activities, for example, to extend half-life and induce host immune responses. New chemical entities could target families of pathogens to create “one-drug-multiple-bug” approaches to replace “one-drug-one-bug” approaches. Traditional vaccine platforms have a history of clinical validation and scaled production capacity. Emerging nucleic acid–based vaccine systems have promise for generating a candidate upon availability of a genomic sequence. Several factors must be considered to rapidly build and advance such a pipeline. Arguably the most critical factor is to incentivize all development groups and encourage aggressive competition. Public sector and philanthropic financing should address the cost of research, clinical trials, manufacturing, and supply agreements, and such financing should be available at the earliest possible part of the product development process. This is essential to overcome developers' decision to avoid product development because of lack of a clear revenue model. This financing, in turn, could stimulate the levels of investment and activity from the private sector observed in COVID-19, including public-private partnerships to advance candidates. A fundamental biological understanding of coronaviruses existed prior to COVID-19 and is necessary to drive product development, but a similar biological understanding needs to be improved for many global infectious diseases ([ 11 ][11]). While under development for COVID-19, predictive, validated preclinical assays, animal models, and human challenge models for infectious diseases would provide faster, cost-efficient methods to eliminate candidates earlier in the development cycle ([ 12 ][12], [ 13 ][13]). Moreover, implementing high-quality, decentralized clinical trials and using existing clinical trial networks could reduce the need for each developer to create complex multicountry clinical trial processes and infrastructure while still maintaining consistent evaluation methods ([ 14 ][14]). Machine learning could help optimize clinical trial design and identify populations most likely to benefit from a candidate, thereby reducing the large sample sizes currently required for late-stage clinical trials ([ 15 ][15]). Consideration should be given to what accelerated and flexible regulatory processes may be adopted from COVID-19, and which regulatory agencies should serve as benchmark approvals for those diseases that predominantly affect resource-limited settings. The manufacturing supply chain may need to be improved for some technologies facing global constraints. Additionally, access, affordability, and availability will need to be addressed to ensure that innovations reach the populations in greatest need. Implementing this strategy is not without risk, and there are challenges to overcome. Development of predictive models and biomarkers has proved difficult with COVID-19. The risk-benefit assessment for accelerated first-in-human testing during an unfolding pandemic may differ compared to that for endemic pathogens. Global capacity for late-stage clinical trials may initially be reached quickly in resource-limited settings. As seen with hydroxychloroquine, early approvals based on limited evidence can occur with compounds that ultimately demonstrate no benefit. The advanced financing available for COVID-19 candidates partially emerged from country-specific interests and, if repeated, may continue to foster inequitable access to new tools globally. Ultimately, the SARS-CoV-2 product development model may need optimization to realistically achieve success across multiple global infectious diseases. Of the ∼250 therapeutics and vaccines in clinical development for global infectious diseases, ∼30% are for HIV and AIDS ([ 3 ][3]). The innovation in antiretroviral medicines was initially sparked by strong political will coupled with streamlined regulatory processes. Growing demand produced attractive returns from resource-wealthy countries. By contrast, the distinct regulatory pathways and government funding to address the growing problem of resistance to antimicrobial agents (such as antibiotics) could not overcome the lack of a revenue model, thereby bankrupting companies that successfully developed safe and efficacious therapies and curtailing development activities. For the recent outbreak of Zika virus beginning in 2015 in the Americas, the time frame from identification of genomic sequences to advancing a nucleic acid vaccine into phase 1 clinical trials occurred within 4 months; but the threat to high-income countries quickly subsided, resulting in stalled product development programs. After nearly 40 years of continuous outbreaks in Africa, the potential global spread of Ebola became evident during the 2014–2016 outbreak and spurred public-private partnerships that recently achieved approval of two vaccines and one therapeutic mAb combination (with a second, single therapeutic mAb under regulatory review). Resource-limited countries are experiencing combined morbidity and mortality impacts from COVID-19: from the disease itself and from other global infectious diseases, owing, in large part, to diversion of resources. Which candidates in clinical trials for COVID-19 will reach regulatory approval, what limitations may come with licensed candidates, and the success of emerging technology platforms are all unknown. However, COVID-19 forced the world to construct a new product development approach, taking what was previously perceived as impossible and turning it into reality. How to implement this approach to address other global infectious diseases that continue to curtail global economic growth and devastate humanity must now be decided. 1. [↵][16]1. Institute for Health Metrics and Evaluation , Global Burden of Disease Study 2019; . 2. [↵][17]1. Policy Cures Research , COVID-19 R&D Tracker Update: 6 August 2020; . 3. [↵][18]1. Policy Cures Research , Neglected Diseases R&D Pipeline Tracker—August 2019; . 4. [↵][19]1. D. E. Gordon et al ., Nature 583, 459 (2020). [OpenUrl][20][CrossRef][21][PubMed][22] 5. [↵][23]1. M. Marovich, 2. J. R. Mascola, 3. M. S. Cohen , JAMA 324, 131 (2020). [OpenUrl][24][CrossRef][25][PubMed][26] 6. [↵][27]1. B. S. Graham , Science 368, 945 (2020). [OpenUrl][28][Abstract/FREE Full Text][29] 7. [↵][30]1. L. Corey, 2. J. R. Mascola, 3. A. S. Fauci, 4. F. S. Collins , Science 368, 948 (2020). [OpenUrl][31][Abstract/FREE Full Text][32] 8. [↵][33]1. J. L. Wilson et al ., Sci. Transl. Med. 12, eaax2550 (2020). 9. [↵][34]1. M. Slaoui, 2. M. Hepburn, , N. Engl. J. Med. 383, 1701 (2020). [OpenUrl][35] 10. [↵][36]1. R. Rappuoli, 2. S. Black, 3. D. E. Bloom , Sci. Transl. Med. 11, eaaw2888 (2019). [OpenUrl][37][FREE Full Text][38] 11. [↵][39]1. M. De Rycker, 2. B. Baragaña, 3. S. L. Duce, 4. I. H. Gilbert , Nature 559, 498 (2018). [OpenUrl][40][CrossRef][41] 12. [↵][42]1. J. Cohen , Science 368, 221 (2020). [OpenUrl][43][Abstract/FREE Full Text][44] 13. [↵][45]1. N. Eyal, 2. M. Lipsitch, 3. P. G. Smith , J. Infect. Dis. 221, 1752 (2020). [OpenUrl][46][PubMed][22] 14. [↵][47]1. COVID-19 Clinical Research Coalition , Lancet 395, 1322 (2020). [OpenUrl][48][PubMed][22] 15. [↵][49]1. W. R. Zame et al ., Stat. Biopharm. Res. 10.1080/19466315.2020.1797867 (2020). Acknowledgments: Thanks to D. Gollaher, B. Hubby, M. Kamarck, I. Pleasure, S. Shome, H. W. Virgin, C. Wells, and G. Yamey for their insightful comments. R.G. is an employee and owns shares of Vir Biotechnology, Inc. The author's opinions expressed in this article do not necessarily reflect Vir's official policy. [1]: #ref-1 [2]: #ref-2 [3]: #ref-3 [4]: #ref-4 [5]: #ref-5 [6]: #ref-6 [7]: #ref-7 [8]: #ref-8 [9]: #ref-9 [10]: #ref-10 [11]: #ref-11 [12]: #ref-12 [13]: #ref-13 [14]: #ref-14 [15]: #ref-15 [16]: #xref-ref-1-1 "View reference 1 in text" [17]: #xref-ref-2-1 "View reference 2 in text" [18]: #xref-ref-3-1 "View reference 3 in text" [19]: #xref-ref-4-1 "View reference 4 in text" [20]: {openurl}?query=rft.jtitle%253DNature%26rft.volume%253D583%26rft.spage%253D459%26rft_id%253Dinfo%253Adoi%252F10.1038%252Fs41586-020-2286-9%26rft_id%253Dinfo%253Apmid%252Fhttp%253A%252F%252Fwww.n%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [21]: /lookup/external-ref?access_num=10.1038/s41586-020-2286-9&link_type=DOI [22]: /lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fsci%2F370%2F6519%2F913.atom [23]: #xref-ref-5-1 "View reference 5 in text" [24]: {openurl}?query=rft.jtitle%253DJAMA%26rft.volume%253D324%26rft.spage%253D131%26rft_id%253Dinfo%253Adoi%252F10.1001%252Fjama.2020.10245%26rft_id%253Dinfo%253Apmid%252F32539093%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [25]: /lookup/external-ref?access_num=10.1001/jama.2020.10245&link_type=DOI [26]: /lookup/external-ref?access_num=32539093&link_type=MED&atom=%2Fsci%2F370%2F6519%2F913.atom [27]: #xref-ref-6-1 "View reference 6 in text" [28]: {openurl}?query=rft.jtitle%253DScience%26rft_id%253Dinfo%253Adoi%252F10.1126%252Fscience.abb8923%26rft_id%253Dinfo%253Apmid%252F32385100%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [29]: /lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEyOiIzNjgvNjQ5NC85NDUiO3M6NDoiYXRvbSI7czoyMjoiL3NjaS8zNzAvNjUxOS85MTMuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9 [30]: #xref-ref-7-1 "View reference 7 in text" [31]: {openurl}?query=rft.jtitle%253DScience%26rft_id%253Dinfo%253Adoi%252F10.1126%252Fscience.abc5312%26rft_id%253Dinfo%253Apmid%252F32393526%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [32]: /lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEyOiIzNjgvNjQ5NC85NDgiO3M6NDoiYXRvbSI7czoyMjoiL3NjaS8zNzAvNjUxOS85MTMuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9 [33]: #xref-ref-8-1 "View reference 8 in text" [34]: #xref-ref-9-1 "View reference 9 in text" [35]: {openurl}?query=rft.jtitle%253DN.%2BEngl.%2BJ.%2BMed%26rft.volume%253D383%26rft.spage%253D1701%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [36]: #xref-ref-10-1 "View reference 10 in text" [37]: {openurl}?query=rft.jtitle%253DScience%2BTranslational%2BMedicine%26rft.stitle%253DSci%2BTransl%2BMed%26rft.aulast%253DRappuoli%26rft.auinit1%253DR.%26rft.volume%253D11%26rft.issue%253D497%26rft.spage%253Deaaw2888%26rft.epage%253Deaaw2888%26rft.atitle%253DVaccines%2Band%2Bglobal%2Bhealth%253A%2BIn%2Bsearch%2Bof%2Ba%2Bsustainable%2Bmodel%2Bfor%2Bvaccine%2Bdevelopment%2Band%2Bdelivery%26rft_id%253Dinfo%253Adoi%252F10.1126%252Fscitranslmed.aaw2888%26rft_id%253Dinfo%253Apmid%252F31217336%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [38]: /lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6MTE6InNjaXRyYW5zbWVkIjtzOjU6InJlc2lkIjtzOjE1OiIxMS80OTcvZWFhdzI4ODgiO3M6NDoiYXRvbSI7czoyMjoiL3NjaS8zNzAvNjUxOS85MTMuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9 [39]: #xref-ref-11-1 "View reference 11 in text" [40]: {openurl}?query=rft.jtitle%253DNature%26rft.volume%253D559%26rft.spage%253D498%26rft_id%253Dinfo%253Adoi%252F10.1038%252Fs41586-018-0327-4%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [41]: /lookup/external-ref?access_num=10.1038/s41586-018-0327-4&link_type=DOI [42]: #xref-ref-12-1 "View reference 12 in text" [43]: {openurl}?query=rft.jtitle%253DScience%26rft.stitle%253DScience%26rft.aulast%253DCohen%26rft.auinit1%253DJ.%26rft.volume%253D368%26rft.issue%253D6488%26rft.spage%253D221%26rft.epage%253D222%26rft.atitle%253DFrom%2Bmice%2Bto%2Bmonkeys%252C%2Banimals%2Bstudied%2Bfor%2Bcoronavirus%2Banswers%26rft_id%253Dinfo%253Adoi%252F10.1126%252Fscience.368.6488.221%26rft_id%253Dinfo%253Apmid%252F32299928%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [44]: /lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEyOiIzNjgvNjQ4OC8yMjEiO3M6NDoiYXRvbSI7czoyMjoiL3NjaS8zNzAvNjUxOS85MTMuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9 [45]: #xref-ref-13-1 "View reference 13 in text" [46]: {openurl}?query=rft.jtitle%253DJ.%2BInfect.%2BDis%26rft.volume%253D221%26rft.spage%253D1752%26rft_id%253Dinfo%253Apmid%252Fhttp%253A%252F%252Fwww.n%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [47]: #xref-ref-14-1 "View reference 14 in text" [48]: {openurl}?query=rft.jtitle%253DLancet%26rft.volume%253D395%26rft.spage%253D1322%26rft_id%253Dinfo%253Apmid%252Fhttp%253A%252F%252Fwww.n%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [49]: #xref-ref-15-1 "View reference 15 in text"

No conference on artificial intelligence (AI), machine learning or robotics would be complete without its fair share of technologists, programmers and engineers. But scan the list of attendees at the 2020 Rise of AI Summit, a hybrid (digital and physical) event this week in Berlin (November 17-18, 2020) and the number of people from health insurance companies, banks and venture capitalists is astonishing. As one of the founders of the event, CEO of Asgard Capital, Fabian Westerheide, said in his opening remarks on "The Next Decade of AI, we are in a'renaissance' of the technology." Westerheide says we're seeing a "refurbishment of ideas from the 1960s, 70s and 80s," combined with the amount of data we have now and today's processing power. He calls it "old ideas, new execution, and new capital."

The artificial intelligence Maithra Raghu studies at Google Brain doesn't have a bedside manner. But she's betting it can still help restore a deeply human, disappearing aspect of modern medicine: personal connection. In a health care system flooded with paperwork and patient data, Raghu sees a natural place for neural networks, which analyze vast amounts of information to find patterns that escape the human eye and use them to churn out diagnoses or health care predictions. To her, the technology could prove to be a powerful tool for processing data that can spare providers more time to spend with patients one-on-one. "Machine learning isn't a magic tool here," said Raghu, a senior research scientist who was recently named a STAT Wunderkind.

This study introduces a framework for quality control of measured weather data, including anomaly detection, and infilling missing values. Weather data is a fundamental input to building performance simulations, in which anomalous values defect the results while missing data lead to an unexpected termination of the simulation process. Traditionally, infilling missing values in weather data is performed through periodic or linear interpolations. However, when missing values exceed many consecutive hours, the accuracy of traditional methods is subject to debate. This study demonstrates how Neural Networks can increase the accuracy of data imputation when compared to other supervised learning methods. The framework is validated by predicting missing temperature and relative humidity data for an observation site, through a network of nearby weather stations in Milan, Italy. Results show that the proposed method can facilitate real-time building simulations with accurate and rapid quality control.

In 1939, shortly before the German invasion of Poland, a British emissary, Lord Lothian, visited the White House with an unusual request. The United Kingdom was unable to protect the world from the Nazis, Lothian told President Franklin Delano Roosevelt. "Anglo-Saxon civilization" would thus need a new guardian. The scepter was falling from British hands, Lothian explained, and the United States must "snatch it up." Though informally made, it was an extraordinary entreaty.

The Machine learning model classifies lithic assemblages of Eastern Africa to identify major incidents of history. Archaic incidents have shaped the existence of human civilization. It tells about the evolution of humans over centuries. Understanding these incidents helps us to have a peek in the past. Moreover, it helps in analyzing the various processes that led to the development of humans.