Search

Urgent applications like wildfire management and renewable energy generation require precise, localized weather forecasts near the Earth's surface. However, weather forecast products from machine learning or numerical weather models are currently generated on a global regular grid, on which a naive interpolation cannot accurately reflect fine-grained weather patterns close to the ground. In this work, we train a heterogeneous graph neural network (GNN) end-to-end to downscale gridded forecasts to off-grid locations of interest. This multi-modal GNN takes advantage of local historical weather observations (e.g., wind, temperature) to correct the gridded weather forecast at different lead times towards locally accurate forecasts. Each data modality is modeled as a different type of node in the graph. Using message passing, the node at the prediction location aggregates information from its heterogeneous neighbor nodes. Experiments using weather stations across the Northeastern United States show that our model outperforms a range of data-driven and non-data-driven off-grid forecasting methods. Our approach demonstrates how the gap between global large-scale weather models and locally accurate predictions can be bridged to inform localized decision-making.

Triggered by the realization that AI emulators can rival the performance of traditional numerical weather prediction models running on HPC systems, there is now an increasing number of large AI models that address use cases such as forecasting, downscaling, or nowcasting. While the parallel developments in the AI literature focus on foundation models -- models that can be effectively tuned to address multiple, different use cases -- the developments on the weather and climate side largely focus on single-use cases with particular emphasis on mid-range forecasting. We close this gap by introducing Prithvi WxC, a 2.3 billion parameter foundation model developed using 160 variables from the Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2). Prithvi WxC employs an encoder-decoder-based architecture, incorporating concepts from various recent transformer models to effectively capture both regional and global dependencies in the input data. The model has been designed to accommodate large token counts to model weather phenomena in different topologies at fine resolutions. Furthermore, it is trained with a mixed objective that combines the paradigms of masked reconstruction with forecasting. We test the model on a set of challenging downstream tasks namely: Autoregressive rollout forecasting, Downscaling, Gravity wave flux parameterization, and Extreme events estimation. The pretrained model with 2.3 billion parameters, along with the associated fine-tuning workflows, has been publicly released as an open-source contribution via Hugging Face.

Recognizing daily activities with unobtrusive sensors in smart environments enables various healthcare applications. Monitoring how subjects perform activities at home and their changes over time can reveal early symptoms of health issues, such as cognitive decline. Most approaches in this field use deep learning models, which are often seen as black boxes mapping sensor data to activities. However, non-expert users like clinicians need to trust and understand these models' outputs. Thus, eXplainable AI (XAI) methods for Human Activity Recognition have emerged to provide intuitive natural language explanations from these models. Different XAI methods generate different explanations, and their effectiveness is typically evaluated through user surveys, that are often challenging in terms of costs and fairness. This paper proposes an automatic evaluation method using Large Language Models (LLMs) to identify, in a pool of candidates, the best XAI approach for non-expert users. Our preliminary results suggest that LLM evaluation aligns with user surveys., Comment: Accepted for publication at UbiComp / ISWC 2024's XAIforU workshop

The sensor-based recognition of Activities of Daily Living (ADLs) in smart home environments enables several applications in the areas of energy management, safety, well-being, and healthcare. ADLs recognition is typically based on deep learning methods requiring large datasets to be trained. Recently, several studies proved that Large Language Models (LLMs) effectively capture common-sense knowledge about human activities. However, the effectiveness of LLMs for ADLs recognition in smart home environments still deserves to be investigated. In this work, we propose ADL-LLM, a novel LLM-based ADLs recognition system. ADLLLM transforms raw sensor data into textual representations, that are processed by an LLM to perform zero-shot ADLs recognition. Moreover, in the scenario where a small labeled dataset is available, ADL-LLM can also be empowered with few-shot prompting. We evaluated ADL-LLM on two public datasets, showing its effectiveness in this domain., Comment: Currently under review

The phenomenology of hadronic states at high energy is well described in the framework of Quantum Chromodynamics. The theory, well established by now, cannot be applied to the description of quark-antiquark states at low energy unless their degrees of freedom are treated as effective ones, a condition resulting from confinement. Own to the similarities with other quantum many-body systems we have adopted a non-perturbative scheme where quarks are treated as quasiparticles, which further interact between them, allowing meson-like states to be described as collective superposition of quasiparticle-pairs. The results of the calculations, performed in the context of the random phase approximation method, show that this scheme is a suitable one to describe meson states up to energies of the order of couple of GeV.

Human Activity Recognition (HAR) based on the sensors of mobile/wearable devices aims to detect the physical activities performed by humans in their daily lives. Although supervised learning methods are the most effective in this task, their effectiveness is constrained to using a large amount of labeled data during training. While collecting raw unlabeled data can be relatively easy, annotating data is challenging due to costs, intrusiveness, and time constraints. To address these challenges, this paper explores alternative approaches for accurate HAR using a limited amount of labeled data. In particular, we have adapted recent Self-Supervised Learning (SSL) algorithms to the HAR domain and compared their effectiveness. We investigate three state-of-the-art SSL techniques of different families: contrastive, generative, and predictive. Additionally, we evaluate the impact of the underlying neural network on the recognition rate by comparing state-of-the-art CNN and transformer architectures. Our results show that a Masked Auto Encoder (MAE) approach significantly outperforms other SSL approaches, including SimCLR, commonly considered one of the best-performing SSL methods in the HAR domain. The code and the pre-trained SSL models are publicly available for further research and development.

Context-aware Human Activity Recognition (HAR) is a hot research area in mobile computing, and the most effective solutions in the literature are based on supervised deep learning models. However, the actual deployment of these systems is limited by the scarcity of labeled data that is required for training. Neuro-Symbolic AI (NeSy) provides an interesting research direction to mitigate this issue, by infusing common-sense knowledge about human activities and the contexts in which they can be performed into HAR deep learning classifiers. Existing NeSy methods for context-aware HAR rely on knowledge encoded in logic-based models (e.g., ontologies) whose design, implementation, and maintenance to capture new activities and contexts require significant human engineering efforts, technical knowledge, and domain expertise. Recent works show that pre-trained Large Language Models (LLMs) effectively encode common-sense knowledge about human activities. In this work, we propose ContextGPT: a novel prompt engineering approach to retrieve from LLMs common-sense knowledge about the relationship between human activities and the context in which they are performed. Unlike ontologies, ContextGPT requires limited human effort and expertise. An extensive evaluation carried out on two public datasets shows how a NeSy model obtained by infusing common-sense knowledge from ContextGPT is effective in data scarcity scenarios, leading to similar (and sometimes better) recognition rates than logic-based approaches with a fraction of the effort.

Machine learning and deep learning methods have been widely explored in understanding the chaotic behavior of the atmosphere and furthering weather forecasting. There has been increasing interest from technology companies, government institutions, and meteorological agencies in building digital twins of the Earth. Recent approaches using transformers, physics-informed machine learning, and graph neural networks have demonstrated state-of-the-art performance on relatively narrow spatiotemporal scales and specific tasks. With the recent success of generative artificial intelligence (AI) using pre-trained transformers for language modeling and vision with prompt engineering and fine-tuning, we are now moving towards generalizable AI. In particular, we are witnessing the rise of AI foundation models that can perform competitively on multiple domain-specific downstream tasks. Despite this progress, we are still in the nascent stages of a generalizable AI model for global Earth system models, regional climate models, and mesoscale weather models. Here, we review current state-of-the-art AI approaches, primarily from transformer and operator learning literature in the context of meteorology. We provide our perspective on criteria for success towards a family of foundation models for nowcasting and forecasting weather and climate predictions. We also discuss how such models can perform competitively on downstream tasks such as downscaling (super-resolution), identifying conditions conducive to the occurrence of wildfires, and predicting consequential meteorological phenomena across various spatiotemporal scales such as hurricanes and atmospheric rivers. In particular, we examine current AI methodologies and contend they have matured enough to design and implement a weather foundation model., Comment: 44 pages, 1 figure, updated Fig. 1

Storing and streaming high dimensional data for foundation model training became a critical requirement with the rise of foundation models beyond natural language. In this paper we introduce TensorBank, a petabyte scale tensor lakehouse capable of streaming tensors from Cloud Object Store (COS) to GPU memory at wire speed based on complex relational queries. We use Hierarchical Statistical Indices (HSI) for query acceleration. Our architecture allows to directly address tensors on block level using HTTP range reads. Once in GPU memory, data can be transformed using PyTorch transforms. We provide a generic PyTorch dataset type with a corresponding dataset factory translating relational queries and requested transformations as an instance. By making use of the HSI, irrelevant blocks can be skipped without reading them as those indices contain statistics on their content at different hierarchical resolution levels. This is an opinionated architecture powered by open standards and making heavy use of open-source technology. Although, hardened for production use using geospatial-temporal data, this architecture generalizes to other use case like computer vision, computational neuroscience, biological sequence analysis and more.

The use of supervised learning for Human Activity Recognition (HAR) on mobile devices leads to strong classification performances. Such an approach, however, requires large amounts of labeled data, both for the initial training of the models and for their customization on specific clients (whose data often differ greatly from the training data). This is actually impractical to obtain due to the costs, intrusiveness, and time-consuming nature of data annotation. Moreover, even with the help of a significant amount of labeled data, model deployment on heterogeneous clients faces difficulties in generalizing well on unseen data. Other domains, like Computer Vision or Natural Language Processing, have proposed the notion of pre-trained models, leveraging large corpora, to reduce the need for annotated data and better manage heterogeneity. This promising approach has not been implemented in the HAR domain so far because of the lack of public datasets of sufficient size. In this paper, we propose a novel strategy to combine publicly available datasets with the goal of learning a generalized HAR model that can be fine-tuned using a limited amount of labeled data on an unseen target domain. Our experimental evaluation, which includes experimenting with different state-of-the-art neural network architectures, shows that combining public datasets can significantly reduce the number of labeled samples required to achieve satisfactory performance on an unseen target domain., Comment: IEEE SMARTCOMP 2023

Deep Learning models are a standard solution for sensor-based Human Activity Recognition (HAR), but their deployment is often limited by labeled data scarcity and models' opacity. Neuro-Symbolic AI (NeSy) provides an interesting research direction to mitigate these issues by infusing knowledge about context information into HAR deep learning classifiers. However, existing NeSy methods for context-aware HAR require computationally expensive symbolic reasoners during classification, making them less suitable for deployment on resource-constrained devices (e.g., mobile devices). Additionally, NeSy approaches for context-aware HAR have never been evaluated on in-the-wild datasets, and their generalization capabilities in real-world scenarios are questionable. In this work, we propose a novel approach based on a semantic loss function that infuses knowledge constraints in the HAR model during the training phase, avoiding symbolic reasoning during classification. Our results on scripted and in-the-wild datasets show the impact of different semantic loss functions in outperforming a purely data-driven model. We also compare our solution with existing NeSy methods and analyze each approach's strengths and weaknesses. Our semantic loss remains the only NeSy solution that can be deployed as a single DNN without the need for symbolic reasoning modules, reaching recognition rates close (and better in some cases) to existing approaches.

Supervised Deep Learning (DL) models are currently the leading approach for sensor-based Human Activity Recognition (HAR) on wearable and mobile devices. However, training them requires large amounts of labeled data whose collection is often time-consuming, expensive, and error-prone. At the same time, due to the intra- and inter-variability of activity execution, activity models should be personalized for each user. In this work, we propose SelfAct: a novel framework for HAR combining self-supervised and active learning to mitigate these problems. SelfAct leverages a large pool of unlabeled data collected from many users to pre-train through self-supervision a DL model, with the goal of learning a meaningful and efficient latent representation of sensor data. The resulting pre-trained model can be locally used by new users, which will fine-tune it thanks to a novel unsupervised active learning strategy. Our experiments on two publicly available HAR datasets demonstrate that SelfAct achieves results that are close to or even better than the ones of fully supervised approaches with a small number of active learning queries.

Abstract: The interaction of dark matter particles (WIMPs) with the odd-mass $^{73}$Ge and $^{131}$Xe target nuclei ,{ {that is the recoil rates corresponding to the elastic scattering of WIMPs by these nuclei}}, is analysed in the context of the minimal extensions of the SUSY model. The BCS+QRPA technique plus the quasiparticle-phonon coupling scheme is used to describe the nuclear structure part of the calculations. The resulting values for the nuclear spin content of both nuclei are compared to values previously reported in the literature.

Joint bleeding is a common condition for people with hemophilia and, if untreated, can result in hemophilic arthropathy. Ultrasound imaging has recently emerged as an effective tool to diagnose joint recess distension caused by joint bleeding. However, no computer-aided diagnosis tool exists to support the practitioner in the diagnosis process. This paper addresses the problem of automatically detecting the recess and assessing whether it is distended in knee ultrasound images collected in patients with hemophilia. After framing the problem, we propose two different approaches: the first one adopts a one-stage object detection algorithm, while the second one is a multi-task approach with a classification and a detection branch. The experimental evaluation, conducted with $483$ annotated images, shows that the solution based on object detection alone has a balanced accuracy score of $0.74$ with a mean IoU value of $0.66$, while the multi-task approach has a higher balanced accuracy value ($0.78$) at the cost of a slightly lower mean IoU value.

The prediction of properties of the low energy portion of the hadronic spectrum is a challenging task which, up to day, is still tentatively given due to the non-perturbative nature of QCD at low energies. In this paper we are exploring the validity of the SO(4) scheme, as representative of the fundamental QCD structure of meson-like states in the region below 2.5 GeV. We have focussed the attention in the calculation of the energy and width of states of various spin, isospin and parities. The theoretical results are compared, sistematically, to the available experimental information., Comment: 16 pages, 11 figures, 9 tables, 36 references

Dark Matter particles can be detected directly via their elastic scattering with nuclei. Next generation experiments can eventually find physical evidences about dark matter candidates. With this motivation in mind, we have calculated the expected signals of dark matter particles in xenon detectors. The calculations were performed by considering different masses and parameters within the minimal supersymmetric standard model. Since the detectors can also detect neutrinos, we have analyzed the supernova neutrino signal including a sterile neutrino in the formalism. Using this 3+1 scheme, we make predictions for both the normal and inverse mass hierarchy. In order to perform a study of the response of planned direct-detection experiments, to be located in ANDES (Agua Negra Deep Experimental Site), we have calculated the neutrino contributions to the background by taken into account reactor's neutrinos and geoneutrinos at the site of the lab. As a test detector, we take a Xenon1T-like array., Comment: 23 pages,11 figures

Among dark-matter candidates are the WIMPs (Weekly Interacting Massive Particles). Low-threshold detectors could directly detect dark-matter by measuring the energy deposited by the particles. In this work we examine the cross section for the elastic scattering of WIMPs on nucleons, in the spin-dependent and spin-independent channels. WIMPs are taken as neutralinos in the context of the minimal super-symmetric extension of the standard model (MSSM). The dependence of the results with the adopted MSSM parameters is discussed., Comment: 21 pages, 5 figures

Neutrinos play an important role in core-collapse supernova events since they are a key piece to understand the explosion mechanisms. The analysis of the neutrino fluxes can bring answers to neutrino's related problems e.g.: mass hierarchy, spectral splitting, sterile neutrinos, etc. In this work we study the impact of neutrino oscillations and the possible existence of eV sterile neutrinos upon the supernova neutrino flux $F_\nu(E)$. We have calculated the energy distribution of the neutrino flux from a supernova and the total number of events which would be detected in a liquid scintillator. We also present an analysis for the conversion probabilities as a function of the active-sterile neutrino mixing parameters. Finally, we have carried out a statistical analysis to extract values for the mixing parameters of the model., Comment: 20 pages, 11 figures, 74 references

The exploration of the non-perturbative regime of QCD, that is the low-energy portion of the hadron spectrum, requires the adoption of theoretical methods more frequently applied to other, more conventional, quantum many body systems, like the atomic nucleus, solid state systems, etc. In this work we have adopted, as a first step, the well-known BCS method to describe correlations between pairs of quarks and the associated ground state. Going beyond the BCS method would imply the inclusion of correlations by means of the TDA or RPA approximations. Since, we are interested in analyzing the role of constituent quark-pair correlations in the structure of hadrons we are restricted to the use of BCS as said before. The starting Hamiltonian is the effective Coulomb plus linear potential which we have used in previous calculations and performed a two-step approach, firstly by pre-diagonalizing it to built a single particle spectrum, and then, secondly, by applying the BCS transformations to it. Then, we have explored the resulting structure of the low energy meson spectra in terms of quasiparticle degrees of freedom. The dependence of the results upon the parameters which enter in the calculations is explored in detail, at the level of the quasiparticle mean-field approximation., Comment: 18 pages, 10 figures, 33 references

Blazars research is one of the hot topics of contemporary extra-galactic astrophysics. That is because these sources are the most abundant type of extra-galactic gamma-ray sources and are suspected to play a central role in multi-messenger astrophysics. We have used swift_xrtproc, a tool to carry out an accurate spectral and photometric analysis of the Swift-XRT data of all blazars observed by Swift at least 50 times between December 2004 and the end of 2020. We present a database of X-ray spectra, best-fit parameter values, count-rates and flux estimations in several energy bands of over 31,000 X-ray observations and single snapshots of 65 blazars. The results of the X-ray analysis have been combined with other multi-frequency archival data to assemble the broad-band Spectral Energy Distributions (SEDs) and the long-term light-curves of all sources in the sample. Our study shows that large X-ray luminosity variability on different timescales is present in all objects. Spectral changes are also frequently observed with a "harder-when-brighter" or "softer-when-brighter" behavior depending on the SED type of the blazars. The peak energy of the synchrotron component nu_peak in the SED of HBL blazars, estimated from the log-parabolic shape of their X-ray spectra, also exhibits very large changes in the same source, spanning a range of over two orders of magnitude in Mrk421 and Mrk501, the objects with the best data sets in our sample., Comment: Accepted for publication in MNRAS

An impact of climate change is the increase in frequency and intensity of extreme precipitation events. However, confidently predicting the likelihood of extreme precipitation at seasonal scales remains an outstanding challenge. Here, we present an approach to forecasting the quantiles of the maximum daily precipitation in each week up to six months ahead using the temporal fusion transformer (TFT) model. Through experiments in two regions, we compare TFT predictions with those of two baselines: climatology and a calibrated ECMWF SEAS5 ensemble forecast (S5). Our results show that, in terms of quantile risk at six month lead time, the TFT predictions significantly outperform those from S5 and show an overall small improvement compared to climatology. The TFT also responds positively to departures from normal that climatology cannot.

The production of heavy-mass elements due to the rapid neutron-capture mechanism (r-process) is associated with astrophysical scenarios, such as supernovae and neutron-star mergers. In the r-process the capture of neutrons is followed by $\beta$-decays until nuclear stability is reached. A key element in the chain of nuclear weak-decays leading to the production of isotopes may be the change of the parameters controlling the neutrino sector, due to the mixing of active and sterile species. In this work we have addressed this question and calculated $\beta$-decay rates for the nuclei involved in the r-process chains as a function of the neutrino mixing parameters. These rates were then used in the calculation of the abundance of the heavy elements produced in core-collapse supernova and in neutron-star mergers, starting from different initial mass-fraction distributions. The analysis shows that the core-collapse supernova environment contributes with approximately $30\%$ of the total heavy nuclei abundance while the neutron-star merger contributes with about $70\%$ of it. Using available experimental data we have performed a statistical analysis to set limits on the active-sterile neutrino mixing angle and found a best-fit value $\sin^2 2\theta_{14}=0.22$, a value comparable with those found in other studies reported in the literature.

The most effective data-driven methods for human activities recognition (HAR) are based on supervised learning applied to the continuous stream of sensors data. However, these methods perform well on restricted sets of activities in domains for which there is a fully labeled dataset. It is still a challenge to cope with the intra- and inter-variability of activity execution among different subjects in large scale real world deployment. Semi-supervised learning approaches for HAR have been proposed to address the challenge of acquiring the large amount of labeled data that is necessary in realistic settings. However, their centralised architecture incurs in the scalability and privacy problems when the process involves a large number of users. Federated Learning (FL) is a promising paradigm to address these problems. However, the FL methods that have been proposed for HAR assume that the participating users can always obtain labels to train their local models. In this work, we propose FedHAR: a novel hybrid method for HAR that combines semi-supervised and federated learning. Indeed, FedHAR combines active learning and label propagation to semi-automatically annotate the local streams of unlabeled sensor data, and it relies on FL to build a global activity model in a scalable and privacy-aware fashion. FedHAR also includes a transfer learning strategy to personalize the global model on each user. We evaluated our method on two public datasets, showing that FedHAR reaches recognition rates and personalization capabilities similar to state-of-the-art FL supervised approaches. As a major advantage, FedHAR only requires a very limited number of annotated data to populate a pre-trained model and a small number of active learning questions that quickly decrease while using the system, leading to an effective and scalable solution for the data scarcity problem of HAR.

Extreme weather events have an enormous impact on society and are expected to become more frequent and severe with climate change. In this context, resilience planning becomes crucial for risk mitigation and coping with these extreme events. Machine learning techniques can play a critical role in resilience planning through the generation of realistic extreme weather event scenarios that can be used to evaluate possible mitigation actions. This paper proposes a modular framework that relies on interchangeable components to produce extreme weather event scenarios. We discuss possible alternatives for each of the components and show initial results comparing two approaches on the task of generating precipitation scenarios.

Machine Learning (ML) has already fundamentally changed several businesses. More recently, it has also been profoundly impacting the computational science and engineering domains, like geoscience, climate science, and health science. In these domains, users need to perform comprehensive data analyses combining scientific data and ML models to provide for critical requirements, such as reproducibility, model explainability, and experiment data understanding. However, scientific ML is multidisciplinary, heterogeneous, and affected by the physical constraints of the domain, making such analyses even more challenging. In this work, we leverage workflow provenance techniques to build a holistic view to support the lifecycle of scientific ML. We contribute with (i) characterization of the lifecycle and taxonomy for data analyses; (ii) design principles to build this view, with a W3C PROV compliant data representation and a reference system architecture; and (iii) lessons learned after an evaluation in an Oil & Gas case using an HPC cluster with 393 nodes and 946 GPUs. The experiments show that the principles enable queries that integrate domain semantics with ML models while keeping low overhead (<1%), high scalability, and an order of magnitude of query acceleration under certain workloads against without our representation., Comment: 21 pages, 10 figures, text overlap with arXiv:1910.04223, a workshop paper being extended in this journal paper