Discover and publish cutting edge, open research.

Nick K. Jones1,2*, Lucy Rivett1,2*, Chris Workman3, Mark Ferris3, Ashley Shaw1, Cambridge COVID-19 Collaboration1,4, Paul J. Lehner1,4, Rob Howes5, Giles Wright3, Nicholas J. Matheson1,4,6¶, Michael P. Weekes1,7¶1 Cambridge University NHS Hospitals Foundation Trust, Cambridge, UK2 Clinical Microbiology & Public Health Laboratory, Public Health England, Cambridge, UK3 Occupational Health and Wellbeing, Cambridge Biomedical Campus, Cambridge, UK4 Cambridge Institute of Therapeutic Immunology & Infectious Disease, University of Cambridge, Cambridge, UK5 Cambridge COVID-19 Testing Centre and AstraZeneca, Anne Mclaren Building, Cambridge, UK6 NHS Blood and Transplant, Cambridge, UK7 Cambridge Institute for Medical Research, University of Cambridge, Cambridge, UK*Joint first authorship¶Joint last authorshipCorrespondence: [email protected] UK has initiated mass COVID-19 immunisation, with healthcare workers (HCWs) given early priority because of the potential for workplace exposure and risk of onward transmission to patients. The UK’s Joint Committee on Vaccination and Immunisation has recommended maximising the number of people vaccinated with first doses at the expense of early booster vaccinations, based on single dose efficacy against symptomatic COVID-19 disease.1-3At the time of writing, three COVID-19 vaccines have been granted emergency use authorisation in the UK, including the BNT162b2 mRNA COVID-19 vaccine (Pfizer-BioNTech). A vital outstanding question is whether this vaccine prevents or promotes asymptomatic SARS-CoV-2 infection, rather than symptomatic COVID-19 disease, because sub-clinical infection following vaccination could continue to drive transmission. This is especially important because many UK HCWs have received this vaccine, and nosocomial COVID-19 infection has been a persistent problem.Through the implementation of a 24 h-turnaround PCR-based comprehensive HCW screening programme at Cambridge University Hospitals NHS Foundation Trust (CUHNFT), we previously demonstrated the frequent presence of pauci- and asymptomatic infection amongst HCWs during the UK’s first wave of the COVID-19 pandemic.4 Here, we evaluate the effect of first-dose BNT162b2 vaccination on test positivity rates and cycle threshold (Ct) values in the asymptomatic arm of our programme, which now offers weekly screening to all staff.Vaccination of HCWs at CUHNFT began on 8th December 2020, with mass vaccination from 8th January 2021. Here, we analyse data from the two weeks spanning 18thto 31st January 2021, during which: (a) the prevalence of COVID-19 amongst HCWs remained approximately constant; and (b) we screened comparable numbers of vaccinated and unvaccinated HCWs. Over this period, 4,408 (week 1) and 4,411 (week 2) PCR tests were performed from individuals reporting well to work. We stratified HCWs <12 days or > 12 days post-vaccination because this was the point at which protection against symptomatic infection began to appear in phase III clinical trial.226/3,252 (0·80%) tests from unvaccinated HCWs were positive (Ct<36), compared to 13/3,535 (0·37%) from HCWs <12 days post-vaccination and 4/1,989 (0·20%) tests from HCWs ≥12 days post-vaccination (p=0·023 and p=0·004, respectively; Fisher’s exact test, Figure). This suggests a four-fold decrease in the risk of asymptomatic SARS-CoV-2 infection amongst HCWs ≥12 days post-vaccination, compared to unvaccinated HCWs, with an intermediate effect amongst HCWs <12 days post-vaccination.A marked reduction in infections was also seen when analyses were repeated with: (a) inclusion of HCWs testing positive through both the symptomatic and asymptomatic arms of the programme (56/3,282 (1·71%) unvaccinated vs 8/1,997 (0·40%) ≥12 days post-vaccination, 4·3-fold reduction, p=0·00001); (b) inclusion of PCR tests which were positive at the limit of detection (Ct>36, 42/3,268 (1·29%) vs 15/2,000 (0·75%), 1·7-fold reduction, p=0·075); and (c) extension of the period of analysis to include six weeks from December 28th to February 7th 2021 (113/14,083 (0·80%) vs 5/4,872 (0·10%), 7·8-fold reduction, p=1x10-9). In addition, the median Ct value of positive tests showed a non-significant trend towards increase between unvaccinated HCWs and HCWs > 12 days post-vaccination (23·3 to 30·3, Figure), suggesting that samples from vaccinated individuals had lower viral loads.We therefore provide real-world evidence for a high level of protection against asymptomatic SARS-CoV-2 infection after a single dose of BNT162b2 vaccine, at a time of predominant transmission of the UK COVID-19 variant of concern 202012/01 (lineage B.1.1.7), and amongst a population with a relatively low frequency of prior infection (7.2% antibody positive).5This work was funded by a Wellcome Senior Clinical Research Fellowship to MPW (108070/Z/15/Z), a Wellcome Principal Research Fellowship to PJL (210688/Z/18/Z), and an MRC Clinician Scientist Fellowship (MR/P008801/1) and NHSBT workpackage (WPA15-02) to NJM. Funding was also received from Addenbrooke’s Charitable Trust and the Cambridge Biomedical Research Centre. We also acknowledge contributions from all staff at CUHNFT Occupational Health and Wellbeing and the Cambridge COVID-19 Testing Centre.

Scene flow tracks the three-dimensional (3D) motion of each point in adjacent point clouds. It provides fundamental 3D motion perception for autonomous driving and server robot. Although the Red Green Blue Depth (RGBD) camera or Light Detection and Ranging (LiDAR) capture discrete 3D points in space, the objects and motions usually are continuous in the macro world. That is, the objects keep themselves consistent as they flow from the current frame to the next frame. Based on this insight, the Generative Adversarial Networks (GAN) is utilized to self-learn 3D scene flow with no need for ground truth. The fake point cloud of the second frame is synthesized from the predicted scene flow and the point cloud of the first frame. The adversarial training of the generator and discriminator is realized through synthesizing indistinguishable fake point cloud and discriminating the real point cloud and the synthesized fake point cloud. The experiments on Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) scene flow dataset show that our method realizes promising results without ground truth. Just as human, the proposed method can identify the similar local structures of two adjacent frames even without knowing the ground truth scene flow. Then, the local correspondence can be correctly estimated, and further the scene flow is correctly estimated.   Corresponding author(s) Email: [email protected]

Leukocyte differential test is a widely performed clinical procedure for screening infectious diseases. Existing hematology analyzers require labor-intensive work and a panel of expensive reagents. Here we report an artificial-intelligence enabled reagent-free imaging hematology analyzer (AIRFIHA) modality that can accurately classify subpopulations of leukocytes with minimal sample preparation. AIRFIHA is realized through training a two-step residual neural network using label-free images of isolated leukocytes acquired from a custom-built quantitative phase microscope. By leveraging the rich information contained in quantitative phase images, we not only achieved high accuracy in differentiating B and T lymphocytes, but also classified CD4 and CD8 cells, therefore outperforming the classification accuracy of most current hematology analyzers. We validated the performance of AIRFIHA in a randomly selected test set and cross-validated it across all blood donors. Owing to its easy operation, low cost, and accurate discerning capability of complex leukocyte subpopulations, we envision AIRFIHA is clinically translatable and can also be deployed in resource-limited settings, e.g., during pandemic situations for the rapid screening of infectious diseases.  Corresponding author(s) Email:    [email protected],  [email protected]

Tam Hunt [1], Jonathan SchoolerUniversity of California Santa Barbara Synchronization, harmonization, vibrations, or simply resonance in its most general sense seems to have an integral relationship with consciousness itself. One of the possible “neural correlates of consciousness” in mammalian brains is a combination of gamma, beta and theta synchrony. More broadly, we see similar kinds of resonance patterns in living and non-living structures of many types. What clues can resonance provide about the nature of consciousness more generally? This paper provides an overview of resonating structures in the fields of neuroscience, biology and physics and attempts to coalesce these data into a solution to what we see as the “easy part” of the Hard Problem, which is generally known as the “combination problem” or the “binding problem.” The combination problem asks: how do micro-conscious entities combine into a higher-level macro-consciousness? The proposed solution in the context of mammalian consciousness suggests that a shared resonance is what allows different parts of the brain to achieve a phase transition in the speed and bandwidth of information flows between the constituent parts. This phase transition allows for richer varieties of consciousness to arise, with the character and content of that consciousness in each moment determined by the particular set of constituent neurons. We also offer more general insights into the ontology of consciousness and suggest that consciousness manifests as a relatively smooth continuum of increasing richness in all physical processes, distinguishing our view from emergentist materialism. We refer to this approach as a (general) resonance theory of consciousness and offer some responses to Chalmers’ questions about the different kinds of “combination problem.”  At the heart of the universe is a steady, insistent beat: the sound of cycles in sync…. [T]hese feats of synchrony occur spontaneously, almost as if nature has an eerie yearning for order. Steven Strogatz, Sync: How Order Emerges From Chaos in the Universe, Nature and Daily Life (2003) If you want to find the secrets of the universe, think in terms of energy, frequency and vibration.Nikola Tesla (1942) I.               Introduction Is there an “easy part” and a “hard part” to the Hard Problem of consciousness? In this paper, we suggest that there is. The harder part is arriving at a philosophical position with respect to the relationship of matter and mind. This paper is about the “easy part” of the Hard Problem but we address the “hard part” briefly in this introduction.  We have both arrived, after much deliberation, at the position of panpsychism or panexperientialism (all matter has at least some associated mind/experience and vice versa). This is the view that all things and processes have both mental and physical aspects. Matter and mind are two sides of the same coin.  Panpsychism is one of many possible approaches that addresses the “hard part” of the Hard Problem. We adopt this position for all the reasons various authors have listed (Chalmers 1996, Griffin 1997, Hunt 2011, Goff 2017). This first step is particularly powerful if we adopt the Whiteheadian version of panpsychism (Whitehead 1929).  Reaching a position on this fundamental question of how mind relates to matter must be based on a “weight of plausibility” approach, rather than on definitive evidence, because establishing definitive evidence with respect to the presence of mind/experience is difficult. We must generally rely on examining various “behavioral correlates of consciousness” in judging whether entities other than ourselves are conscious – even with respect to other humans—since the only consciousness we can know with certainty is our own. Positing that matter and mind are two sides of the same coin explains the problem of consciousness insofar as it avoids the problems of emergence because under this approach consciousness doesn’t emerge. Consciousness is, rather, always present, at some level, even in the simplest of processes, but it “complexifies” as matter complexifies, and vice versa. Consciousness starts very simple and becomes more complex and rich under the right conditions, which in our proposed framework rely on resonance mechanisms. Matter and mind are two sides of the coin. Neither is primary; they are coequal.  We acknowledge the challenges of adopting this perspective, but encourage readers to consider the many compelling reasons to consider it that are reviewed elsewhere (Chalmers 1996, Griffin 1998, Hunt 2011, Goff 2017, Schooler, Schooler, & Hunt, 2011; Schooler, 2015).  Taking a position on the overarching ontology is the first step in addressing the Hard Problem. But this leads to the related questions: at what level of organization does consciousness reside in any particular process? Is a rock conscious? A chair? An ant? A bacterium? Or are only the smaller constituents, such as atoms or molecules, of these entities conscious? And if there is some degree of consciousness even in atoms and molecules, as panpsychism suggests (albeit of a very rudimentary nature, an important point to remember), how do these micro-conscious entities combine into the higher-level and obvious consciousness we witness in entities like humans and other mammals?  This set of questions is known as the “combination problem,” another now-classic problem in the philosophy of mind, and is what we describe here as the “easy part” of the Hard Problem. Our characterization of this part of the problem as “easy”[2] is, of course, more than a little tongue in cheek. The authors have discussed frequently with each other what part of the Hard Problem should be labeled the easier part and which the harder part. Regardless of the labels we choose, however, this paper focuses on our suggested solution to the combination problem.  Various solutions to the combination problem have been proposed but none have gained widespread acceptance. This paper further elaborates a proposed solution to the combination problem that we first described in Hunt 2011 and Schooler, Hunt, and Schooler 2011. The proposed solution rests on the idea of resonance, a shared vibratory frequency, which can also be called synchrony or field coherence. We will generally use resonance and “sync,” short for synchrony, interchangeably in this paper. We describe the approach as a general resonance theory of consciousness or just “general resonance theory” (GRT). GRT is a field theory of consciousness wherein the various specific fields associated with matter and energy are the seat of conscious awareness.  A summary of our approach appears in Appendix 1.  All things in our universe are constantly in motion, in process. Even objects that appear to be stationary are in fact vibrating, oscillating, resonating, at specific frequencies. So all things are actually processes. Resonance is a specific type of motion, characterized by synchronized oscillation between two states.  An interesting phenomenon occurs when different vibrating processes come into proximity: they will often start vibrating together at the same frequency. They “sync up,” sometimes in ways that can seem mysterious, and allow for richer and faster information and energy flows (Figure 1 offers a schematic). Examining this phenomenon leads to potentially deep insights about the nature of consciousness in both the human/mammalian context but also at a deeper ontological level.

Cannabis sativa is an extraordinarily versatile species. Hemp and its cousin marijuana, both C. sativa, have been used for millennia as a source of fibre, oil and for medicinal, spiritual and recreational purposes. Because the consumption of Cannabis can have psychoactive effects, the plant has been widely banned throughout the last century. In the past decade, evidence of its medicinal properties did lead to the relaxation of legislation in many countries around the world. Consequently, the genetics and development of Cannabis as well as Cannabis-derived products are the subject of renewed attention.Here, we review the biology of C. sativa, including recent insights from taxonomy, morphology and genomics, with an emphasis on the genetics of cannabinoid synthesis. Because the female Cannabis flower is of special interest as the site of cannabinoid synthesis, we explore flower development, flowering time well as the species’ unique sex determination system in detail. Furthermore, we outline the tremendous medicinal, engineering, and environmental opportunities that Cannabis bears. Together, the picture emerges that our understanding of Cannabis biology currently progresses at an unusual speed. A future challenge will be to preserve the multi-purpose nature of Cannabis, and to harness its medicinal properties and sustainability advantages simultaneously.

Quantum chemistry must evolve if it wants to fully leverage the benefits of the internet age, where the world wide web offers a vast tapestry of tools that enable users to communicate and interact with complex data at the speed and convenience of a button press. The Open Chemistry project has developed an open source framework that offers an end-to-end solution for producing, sharing, and visualizing quantum chemical data interactively on the web using an array of modern tools and approaches. These tools build on some of the best open source community projects such as Jupyter for interactive online notebooks, coupled with 3D accelerated visualization, state-of-the-art computational chemistry codes including NWChem and Psi4 and emerging machine learning and data mining tools such as ChemML and ANI. They offer flexible formats to import and export data, along with approaches to compare computational and experimental data.

This paper responds to one by Graham Martin and colleagues, who offered a critique of my previous publications on masks for the lay public in the Covid-19 pandemic. I address their charges that my co-authors and I had misapplied the precautionary principle; drawn conclusions that were not supported by empirical research; and failed to take account of potential harms. But before that, I remind Martin et al that the evidence on mask wearing goes beyond the contested trials and observational studies they place centre stage. I set out some key findings from basic science, epidemiology, mathematical modelling, case studies and natural experiments, and use this rich and diverse body of evidence as the backdrop for my rebuttal of their narrowly-framed objections. I challenge my critics’ apparent assumption that a particular kind of systematic review should be valorised over narrative and real-world evidence, since stories are crucial to both our scientific understanding and our moral imagination. I conclude by thanking my academic adversaries for the intellectual sparring match, but exhort them to remember our professional accountability to a society in crisis. It is time to lay straw men to rest and engage, scientifically and morally, with the dreadful tragedy that is unfolding across the world.

This Supporting information includes interactive plots, videos, and data captured while performing evaluation and validation experiments for our paper. 

The construct of wellbeing has been criticised as a neoliberal construction of western individualism that ignores wider systemic issues including increasing burden of chronic disease, widening inequality, concerns over environmental degradation and anthropogenic climate change. While these criticisms overlook recent developments, there remains a need for biopsychosocial models that extend theoretical grounding beyond individual wellbeing, incorporating overlapping contextual issues relating to community and environment. Our first GENIAL model \cite{Kemp_2017} provided a more expansive view of pathways to longevity in the context of individual health and wellbeing, emphasising bidirectional links to positive social ties and the impact of sociocultural factors. In this paper, we build on these ideas and propose GENIAL 2.0, focusing on intersecting individual-community-environmental contributions to health and wellbeing, and laying an evidence-based, theoretical framework on which future research and innovative therapeutic innovations could be based. We suggest that our transdisciplinary model of wellbeing - focusing on individual, community and environmental contributions to personal wellbeing - will help to move the research field forward. In reconceptualising wellbeing, GENIAL 2.0 bridges the gap between psychological science and population health health systems, and presents opportunities for enhancing the health and wellbeing of people living with chronic conditions. Implications for future generations including the very survival of our species are discussed.  

Floral plantings are promoted to foster ecological intensification of agriculture through provisioning of ecosystem services. However, a comprehensive assessment of the effectiveness of different floral plantings, their characteristics and consequences for crop yield across global regions is lacking. Here we quantified the impacts of flower strips and hedgerows on pest control and pollination services in adjacent crops using a global dataset of 529 sites. Flower strips, but not hedgerows, enhanced pest control services in adjacent fields by 16% on average. However, effects on crop pollination and yield were more variable. Our synthesis identifies several important drivers of variability in effectiveness of plantings: pollination services declined exponentially with distance from plantings, and perennial and older flower strips with higher flowering plant diversity enhanced pollination more effectively. These findings provide promising pathways to optimize floral plantings to more effectively contribute to ecosystem service delivery and ecological intensification of agriculture in the future.

Researchers' ability to accurately screen fossil and subfossil specimens for preservation of DNA and protein sequences remains limited. Thermal exposure and geologic age are usable proxies for sequence preservation on a broad scale but are of nominal use for specimens of similar depositional environments. Cell and tissue biomolecular histology is thus proposed as a novel proxy for determining sequence preservation potential of ancient specimens with improved accuracy. Biomolecular histology as a proxy is hypothesized to elucidate why fossils/subfossils of some depositional environments preserve sequences while others do not and to facilitate selection of ancient specimens for use in molecular studies.

IntroductionConsistent with World Health Organization (WHO) advice [1], UK Infection Protection Control guidance recommends that healthcare workers (HCWs) caring for patients with coronavirus disease 2019 (COVID-19) should use fluid resistant surgical masks type IIR (FRSMs) as respiratory protective equipment (RPE), unless aerosol generating procedures (AGPs) are being undertaken or are likely, when a filtering face piece 3 (FFP3) respirator should be used [2]. In a recent update, an FFP3 respirator is recommended if “an unacceptable risk of transmission remains following rigorous application of the hierarchy of control” [3]. Conversely, guidance from the Centers for Disease Control and Prevention (CDC) recommends that HCWs caring for patients with COVID-19 should use an N95 or higher level respirator [4]. WHO guidance suggests that a respirator, such as FFP3, may be used for HCWs in the absence of AGPs if availability or cost is not an issue [1].A recent systematic review undertaken for PHE concluded that: “patients with SARS-CoV-2 infection who are breathing, talking or coughing generate both respiratory droplets and aerosols, but FRSM (and where required, eye protection) are considered to provide adequate staff protection” [5]. Nevertheless, FFP3 respirators are more effective in preventing aerosol transmission than FRSMs, and observational data suggests that they may improve protection for HCWs [6]. It has therefore been suggested that respirators should be considered as a means of affording the best available protection [7], and some organisations have decided to provide FFP3 (or equivalent) respirators to HCWs caring for COVID-19 patients, despite a lack of mandate from local or national guidelines [8].Data from the HCW testing programme at Cambridge University Hospitals NHS Foundation Trust (CUHNFT) during the first wave of the UK severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic indicated a higher incidence of infection amongst HCWs caring for patients with COVID-19, compared with those who did not [9]. Subsequent studies have confirmed this observation [10, 11]. This disparity persisted at CUHNFT in December 2020, despite control measures consistent with PHE guidance and audits indicating good compliance. The CUHNFT infection control committee therefore implemented a change of RPE for staff on “red” (COVID-19) wards from FRSMs to FFP3 respirators. In this study, we analyse the incidence of SARS-CoV-2 infection in HCWs before and after this transition.

The San Joaquin Valley of California is a paradox. It is highly productive agriculturally, but highly vulnerable to hydroclimatic shocks. A principal means of improving systemwide water management flexibility is strategic land fallowing. To date, however, land fallowing is sporadic, ephemeral, and unpredictable. Understanding patterns and drivers of land fallowing would allow resource managers to better plan for water allocation to cities, farms, and ecosystems. The inability to accurately anticipate the distribution of fallow land hampers the optimization of resource use, thereby affecting both the agricultural economy and regional water management. Thus, forecasting fallow land would allow for more optimum agricultural productivity and water delivery, particularly to vulnerable communities and ecosystems. To address this issue, this study compares state-of-the-art machine learning models with deep learning methods using a rich dataset that includes satellite images, weather patterns, soil properties, historical land use information, and other pertinent geospatial data. We employ an advanced ensemble of machine learning algorithms, i.e., Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Support Vector Machines (SVM), for forecasting fallow land. Additionally, we investigate the use of Convolutional and Recurrent Neural Networks (like ConvLSTM, CNN-LSTM, and GAN) to take advantage of the spatial and temporal relationships contained in the data. Our preliminary findings suggest that deep learning models, with their capacity to capture intricate spatial and temporal correlations within the data, may perform better than more conventional machine learning techniques. As a result, these deep learning models could help water managers, policymakers, and researchers make more precise forecasts about the distribution of fallow land, leading to more optimal use of water resources and agricultural output. The results of this study can not only offer insights for California water managers and policymakers, but also enable practical plans for more optimal resource allocation, sustainable agricultural land use, and long-term water security. Our approach may also be applicable to other agricultural regions facing similar challenges globally.

Atmospheric AI modeling is increasingly reliant on complex machine learning (ML) techniques and high-dimensional gridded inputs to develop models that achieve high predictive skill. Complex deep learning architectures such as convolutional neural networks and transformers are trained to model highly non-linear atmospheric phenomena such as coastal fog [1], tornadoes [2], and severe hail [3]. The input data is typically in the form of gridded spatial data composed of multiple channels of satellite imagery, numerical weather prediction output, reanalysis products, etc. In many studies, the use of complex architectures and high-dimensional inputs were shown to substantially outperform simpler alternatives. A major challenge when using complex ML techniques is that it is very difficult to understand how the trained model works. The complexity of the model obfuscates the relationship between the input and prediction. It is often of interest to understand a model’s decision-making process. By exposing the model’s behavior, users could verify that the model has learned physically realistic predictive patterns. This information can be used to calibrate trust in the model. The model may have also learned novel patterns within the data that could be used to gain new insights into the atmospheric process. Extracting learned patterns could be used to generate hypotheses for scientific discovery. The rapid adoption of complex ML models and the need to understand how they work has led to the development of a broad class of techniques called eXplainable Artificial Intelligence (XAI). These methods probe the models in various ways to reveal insights into how they work. Correlations among input features can make it challenging to produce meaningful explanations. The gridded spatial data common in atmospheric modeling applications typically have extensive correlation. Spatial autocorrelation is present among the cells of each spatial grid, but autocorrelation may exist across the gridded data volume due to spatial or temporal relationships between adjacent channels. In addition, there may be correlations between distant locations due to teleconnections between them. Correlated input features may cause high variance among the trained models. If grid cells are highly correlated, then the target function that the network is attempting to learn is ill-defined and an infinite number of models can be generated that achieve approximately equal performance. Even assuming a perfect XAI method exists, the attribution reflects only the patterns learned for a given model. It is arbitrary which of the correlated features are used by a given model. This can lead to a misleading understanding of the actual relationship between the input features and target. A potential solution is to group the correlated features before applying XAI. Attribution can be assigned to each group rather than to individual cells. In this case, all the correlated cells will be permuted at the same time to analyze their collective impact on the output. The purpose is to reveal the contribution of each group of related cells toward the model output. Ideally, the explanations are insensitive to the random choice among correlated features learned by the model. Without grouping, the user can be misled to consider a feature as not being related to the target because of the presence of correlated features. With grouping, the explanations should better reveal the learned patterns. Grouping features based on correlation can be challenging. The correlation rarely equals one and the strength of the correlation influences the variance among trained models. Calculating the correlation can be difficult because of partial correlations and fuzzy, continuous boundaries. The choice of groups can greatly influence the explanations. Another challenge is that it is not straight-forward to assess the quantitative accuracy of an XAI technique. This is because there is rarely a ground truth explanation to compare to. If we knew the attribution, we would not need XAI methods. Synthetic benchmarks for analyzing XAI have been proposed as a solution [4]. It is possible to define a non-linear function such that the contribution of each grid cell’s value to the function output can be derived. This attribution map represents the ground truth for comparison the the output of XAI methods that are applied to a model that very closely approximates the hand-crafted function. In this research, we develop a set of benchmarks to investigate the influence of correlated features on the variation in XAI outputs for a set of trained models. We then explore how features can be grouped to reduce the explanation variance so that users have improved insight into the learned patterns.  First, we create a set of very simple mathematical demonstrations that precisely demonstrate the influence of correlated features and how grouping features provides a solution. Using insights from these experiments, we develop a tool for detecting when correlated features are likely to cause misleading explanations. We then create a set of more realistic benchmarks that are based on atmospheric modeling problems such as sea surface temperature and coastal fog prediction. By defining benchmarks with known ground truth explanations, we can analyze various techniques for grouping the grid cells based on their correlations. Based on our findings, we offer recommendations for strategies to group correlated data so that users can better leverage XAI results toward model development and scientific insights. [1] Kamangir, H., Collins, W., Tissot, P., King, S. A., Dinh, H. T. H., Durham, N., & Rizzo, J. (2021). FogNet: A multiscale 3D CNN with double-branch dense block and attention mechanism for fog prediction. Machine Learning with Applications, 5, 100038.[2] Lagerquist, R. (2020). Using Deep Learning to Improve Prediction and Understanding of High-impact Weather.[3] Gagne II, D. J., Haupt, S. E., Nychka, D. W., & Thompson, G. (2019). Interpretable deep learning for spatial analysis of severe hailstorms. Monthly Weather Review, 147(8), 2827-2845.[4] Mamalakis, A., Ebert-Uphoff, I., & Barnes, E. A. (2022). Neural network attribution methods for problems in geoscience: A novel synthetic benchmark dataset. Environmental Data Science, 1, e8.

Transitioning to renewable energy in the distribution grid (DG) is essential for combating climate change and ensuring energy security. However, this transition can introduce grid instability. To combat this, we need improved control capabilities for these energy resources; which requires accurate information on system state variables and distribution grid line parameters. This study presents a way to simultaneously estimate the system state variables, active and reactive power, and the line parameters of the distribution grid without the need for any information on voltage angles. This is achieved by formulating it as a maximum likelihood problem that we solve using the expectation maximization (EM) algorithm, which we adapt to this problem and provide details of a numerically robust implementation. The study uses the modified Distflow model which provides a way to consider line losses in the system and improves system accuracy. The proposed method is demonstrated on the IEEE 37-node test feeder. The proposed study is compared to state-of-the-art, where we achieve a 70% reduction in voltage error and more than 10, 000 times lower error for state variables.

This perspective conveys a first-person account of the harsh realities of experiencing homelessness, including constant dampness, hunger, isolation, aimlessness, uncertainty, and emotional hardship. Individuals of all ages and backgrounds have or will confront housing insecurity and homelessness. While factors like finances, job loss, and lack of social support are key drivers, mental health issues and substance abuse also contribute to many people becoming homeless. Unexpected acts of kindness from strangers provided glimmers of human connection and hope. Seeing patients as individual human beings rather than faceless diagnoses is connected to recognizing the humanity and dignity in all community members, including the homeless. Compassionate action and social measures to provide shelter and uplift those without homes are vital, affirming their value and belonging.

The actual paper presents an in-depth study and experimental development of a class of rotorcraft, named as x-tilt, that features four tilting rotors. Initially, the equations of motion modeling the aerial robot are presented based on the Euler-Lagrange formulation. The model includes the aerodynamic effects induced by the rotorcraft's relative motion and propellers. For control purposes the aforementioned model is split into a nominal model and lumped disturbance terms, the latter encompassing endogenous and exogenous uncertainties. In this vein, the actual work propose a robust navigation strategy targeting a specific performance profile whose problem is formulated through the model predictive control (MPC) framework. To this end, two schemes are proposed, (i) an integral MPC and a (ii) MP sliding-mode Control (MPSMC). Both control schemes are linked to a extended-state Linear Kalman Filter (ES-LKF) that furnishes the states and lumped disturbance estimates. Moreover, a high-fidelity simulation is presented in detail to validate the effectiveness of the proposed controller within a realistic scenario. We finally present the experimental stage to validate the tilting-rotor configuration as well as the integral MPC.

Chlorine radical (Cl•) profoundly impacts atmospheric oxidation capacity, and therefore climate change and air pollution. Chloramines, especially trichloramine (NCl 3), is a potentially important Cl• source, while their formation mechanism remains elusive. Here, we present evidence of aerosol aqueous-phase production of chloramines with observations in Beijing and develop a comprehensive box model which incorporates ~40 aerosol aqueous reactions initiated by Cl 2 and NH 3 and multiple phase transfer processes. The model can well reproduce the diurnal variation of NCl 3. We show that this atmospheric production of chloramines could be ubiquitous globally with the coexistence of aerosol water, Cl 2 , and NH 3 , while NCl 3 formation is mainly influenced by Cl 2 and relative humidity as NH 3 is typically in great excess. The contribution of chloramines (mostly NCl 3) to Cl• production is much higher in clean periods (up to 85%) than polluted periods (~5%) in Beijing, and in clean environments (e.g., ~64% in Canada) than polluted areas (e.g., ~1% in India). Therefore, chloramines can play an increasingly important role in a cleaner future and may partially offset the ongoing efforts on air pollution mitigation. The missing representation of chloramines in current models could significantly underestimate the climate, health and ecological impacts of Cl•.

This paper reviews about the field of data science and the importance of it with the today's rapidly evolving landscape of technology. We start the survey by introducing the fundamental concepts of data science including history and how we collect data and why we need data science for and how we use them effectively to use them for processing, storing, and analyzing such as machine learning, data visualization. And we explore about the different domains we use data science and the challenges, advantage and disadvantages we face while using them. And finally, we are going to discuss about the prospects and implications of data science and how we can use data science to overcome the current and future challenges. The need of careers and the paths of careers are also discussed as they are essential and growing rapidly.

A document by Daniele Visioni. Click on the document to view its contents.

Distinguished from most hyperspectral Anomaly Detection(AD) methods based on trainable parameter networks, the recently proposed method called AETNet eliminates the need for parameter adjustments or retraining on new test scenes by training an anomaly enhancement network on background data with false anomalies. In this letter, we achieve this by proposing a novel training and inference framework that enhances the network's background spectral feature extraction capability without any data augmentation. During training on background data, the complete network is trained using the reverse distillation framework with a spectral feature alignment mechanism to improve the network's background feature expressiveness. For inference, a pruned network is applied, composed solely of components most relevant to expressing features in the spectral dimension. This effectively reduces redundant information, enhancing both inference efficiency and anomaly detection accuracy. Experimental results demonstrate that our method outperforms state-of-theart methods on the HAD100 dataset, striking an optimal balance between detection accuracy and inference speed. Our code is available at https://github.com/cristianoKaKa/FERD.

The application of millimeter-Wave (mmWave) Radar sensors for people monitoring raised a lot of interest in the context of Active Assisted Living (AAL), especially since the processing of Radar signals can provide interesting information about the observed subjects. Correct recognition of the ongoing behavior, however, cannot disregard from detecting where the subject is positioned. Detection approaches, based on Constant False Alarm Rate (CFAR) algorithms, sometimes fail to correctly identify the presence of targets within the observed scenario, especially in complex environments such as indoors. This paper proposes the use of a mmWave Multiple Input Multiple Output (MIMO) Radar in combination with a You Only Look Once (YOLO) neural network-based algorithm for the detection of moving people in indoor environments by processing all the data cube information at the same time. Results are validated through experimental tests which involve subjects walking in linear or random mode, different Radar configurations, and different indoor environments. By exploiting at the same time information such as the angle, Doppler, and range distance of the target, the proposed approach proves to be very effective in the examined scenarios. Experimental results will be discussed in this work to demonstrate the effectiveness of the proposed method.

Whereas robots are and will be increasingly present in all areas of society, robotics can be limited by its narrow perspective, often neglegecting discussions about its own goals, epistemology, ethics, and socio-political impacts. In order to try to adress this issue, this article proposes a holistic analysis of robots and robotics. Taking into these blind spots, raising new challenges. In response, the second part proposes a set of guidelines : a new epistemology, a metaethical framework, an organization method and a template design in order to answer these challenges.

In an era where the drumbeats of technological advancement echo through the corridors of military strategy, my research takes a deep dive into the storied pasts of military legends-General John J. Pershing, General George S. Patton Jr., and General Norman Schwarzkopf-to juxtapose their timeless strategies with the burgeoning field of Artificial Intelligence (AI) in warfare. This comparative analysis, crafted with the meticulousness of a New York Times feature, seeks to unravel the complex tapestry of leadership, tactical innovation, and the human element that defined the battlegrounds of the 20th century, and to critically examine how AI might redefine the very fabric of military operations in the future.