324677 research outputs found
Sort by
On the application of neural networks for structured domains to fMRI data
No full textFunctional Magnetic Resonance Imaging (fMRI) provides spatio-temporal maps of brain activity; however, extracting the rich information they contain is challenging. Traditional approaches use only summary statistics, losing details that might be hidden in the complex temporal dynamics. Deep neural networks are emerging as an apt solution in this context, given their ability to handle vast amounts of structured data. In this paper, we consider two widely studied fMRI datasets: the Human Connectome Project for connectome fingerprinting, and ABIDE for autism classification. We aim to understand how handling the temporal and spatial dimensions could influence the performance of the models and their interpretability. Specifically, we compare neural network models with architectural biases toward temporal, spatial, or combined spatio-temporal features. The results of our analysis show that existing methods exploiting the spatial dimension, or spatio-temporal hybrids, are not competitive with simpler ones considering the temporal dimension only, such as LSTM. Additionally, we propose a contrastive learning approach for connectome fingerprinting, enabling robust individual identification without requiring access to all subjects during training. Our findings suggest that explicit graph modeling of the interaction between brain regions introduces complexity without improving performance, thereby challenging current trends
Polynomial meshes on algebraic sets
No full textPolynomial meshes (called sometimes 'norming sets') allow us to estimate the supremum norm of polynomials on a fixed compact set by the norm on its discrete subset. We give a general construction of polynomial weakly admissible meshes on compact subsets of arbitrary algebraic hypersurfaces in CN+1. They are preimages by a projection of meshes on compacts in CN. The meshes constructed in this way are optimal in some cases. Our method can be useful also for certain algebraic sets of codimension greater than one. To illustrate applications of the obtained theorems, we first give a few examples and finally report some numerical results. In particular, we present numerical tests (implemented in Matlab), concerning the use of such optimal polynomial meshes for interpolation and least-squares approximation, as well as for the evaluation of the corresponding Lebesgue constants
Optimal aggregation of users to form Energy Communities
No full textEnergy Communities are a key strategy for decarbonization. Nevertheless, their diffusion remains limited, partly due to the difficulty of engaging local participants. In particular, stakeholders investing in renewable energy systems lack practical tools to aggregate local consumers and enhance the economic attractiveness of their projects. This paper introduces a novel aggregation approach to determine the optimal number and type of participants in an Energy Community. A two-level evolutionary algorithm is developed to select the most suitable participants from a pool of end-users belonging to the same local area. The methodology is applied to a case study involving a producer investing in a 1 MW photovoltaic system, with the goal of finding consumers with whom to form the Energy Community. The perspectives of the various stakeholders are reflected through dedicated objective functions. In the configuration identified as the most favourable, the Energy Community includes the producer, four commercial end-users, and a small single-shift factory, while residential end-users are excluded. The producer achieves payback in seven years, while consumers benefit from an 8 % reduction in electricity bills. Additionally, 94 % of the energy made available by the photovoltaic system is utilized locally, thereby limiting reverse power flows to the primary transformer
Linking printing layer thickness to performance-critical microstructure and corrosion in LPBF AlSi7Mg
No full textUnlocking the potential of laser powder bed fusion (LPBF) AlSi7Mg for high-performance applications hinges on a deep understanding of its microstructural response to processing parameters. Previous studies investigating the layer thickness influence on LPBF AlSi7Mg have primarily focused on thicker printing layers (30 μm–100 μm), which neglects the potential impact of very thin layers (e.g., 20–30 μm) on the microstructure and overall properties like service stability. This study aims to fill this knowledge gap by systematically examining the effect of printing layer thickness within this thin critical range on the microstructural evolution and subsequent corrosion behavior of LPBF AlSi7Mg (after T6 heat treatment). Our results demonstrated that the 30 μm printing thickness condition consistently exhibited superior microhardness and corrosion resistance. The detailed microstructural and phase formation analysis revealed that layer thickness has a direct impact on cooling rate and resultant element distribution, which can induce the formation of different Fe-bearing phases like π-AlFeMgSi and β-AlFeSi. Along with different phase formations, the grain boundary (GB) and Si-rich phase concentration and distribution significantly influenced performance by disturbing the passivation layer, which provides valuable insights for optimizing LPBF processing parameters, enhancing the reliability of AlSi7Mg components, and advancing the understanding of this critical material for demanding applications
Revealing The Secret Power: How Algorithms Can Influence Content Visibility on Twitter/X
No full textIn recent years, the opaque design and the limited public understanding of
social networks' recommendation algorithms have raised concerns about potential
manipulation of information exposure. Reducing content visibility, aka shadow
banning, may help limit harmful content; however, it can also be used to
suppress dissenting voices. This prompts the need for greater transparency and
a better understanding of this practice.
In this paper, we investigate the presence of visibility alterations through
a large-scale quantitative analysis of two Twitter/X datasets comprising over
40 million tweets from more than 9 million users, focused on discussions
surrounding the Ukraine-Russia conflict and the 2024 US Presidential Elections.
We use view counts to detect patterns of reduced or inflated visibility and
examine how these correlate with user opinions, social roles, and narrative
framings. Our analysis shows that the algorithm systematically penalizes tweets
containing links to external resources, reducing their visibility by up to a
factor of eight, regardless of the ideological stance or source reliability.
Rather, content visibility may be penalized or favored depending on the
specific accounts producing it, as observed when comparing tweets from the Kyiv
Independent and RT.com or tweets by Donald Trump and Kamala Harris. Overall,
our work highlights the importance of transparency in content moderation and
recommendation systems to protect the integrity of public discourse and ensure
equitable access to online platforms
Efficacy of Mediterranean Diet for the prevention of neurological diseases: a systematic review and meta-analysis featured in the Italian National Guidelines "La Dieta Mediterranea"
No full textA patient-derived decellularized extracellular matrix hydrogel as a biomimetic scaffold for advanced 3D colorectal cancer modeling
Full text linkColorectal cancer (CRC) is among the most prevalent cancers globally and is associated with a high mortality rate, particularly in advanced stages. In the realm of drug discovery, the use of innovative and highly translational pre-clinical CRC models is essential. Currently, the most relevant in vitro tumor approaches are three dimensional (3D) models. However, most 3D models of solid tumors are based either on synthetic materials or animal-derived commercial hydrogels, which fail to accurately mimic the biology of native tissues and originate from non-human sources. In contrast, hydrogels derived from human decellularized extracellular matrix (ECM) retain signaling cues from native tissue and represent a bioactive mechanical structure that can foster tumor cell growth in a tissue-specific 3D in vitro environment. Here, we demonstrated that patient-derived decellularized colon ECM can be processed into a hydrogel, producing the CologEM. CologEM formulation process preserved key ECM proteins, such as collagens, glycosaminoglycans and secreted bioactive molecules belonging to the family of cytokine, chemokine, interleukin, growth factors and ECM-remodeling enzyme. CologEM displayed a fibrous ultrastructure with interconnected pores, with notable differences observed between 1 % and 3 % (w/v) CologEM. Both 1 % and 3 % CologEM showed good biocompatibility, with 3 % CologEM demonstrating a higher propensity to induce a mesenchymal phenotype and resistance to antitumor drugs. In conclusion, CologEM is a suitable scaffold for 3D CRC models as it replicates critical characteristics of the tumor microenvironment. This model holds promise for facilitating the discovery and development of chemotropic drugs for cancer treatment
Design and characterization of MASI On-Orbit Servicing gripper trigger system based on pose measurement logic
No full textThis paper presents the development and testing of a contactless pose measurement system for space docking and berthing operations. The system is part of the Modular Androgynous Standard Interface (MASI) and is compatible with the Mechanical Interface for Capture at End-of-Life (MICE). The goal is to create a device that provides ultra-close proximity pose estimations to automatically trigger the MASI capture mechanism during the On-Orbit Servicing (OOS) capturing phase. The system uses nine infrared (IR) reflective proximity sensors to measure five degrees of freedom: three distances and two angles. One main advantage is that the target interface does not require sensors or markers. Moreover, by providing continuous measurements at 50 Hz, it enables real-time adjustments of the relative pose between the servicer and the target, reducing potential failures during docking or berthing operations. A mathematical model describing the relative pose between the sensorized and target interfaces was identified and validated through calibration. The testing was carried out on a hexapod platform to simulate different positions and orientations. The results show a strong and accurate performance throughout the measurement range. The system has reached Technology Readiness Level (TRL) 4. Future work will aim to improve the performance of the system and make the test environment more realistic to increase the TRL
SBV functions in Carnot–Carathéodory spaces
No full textWe introduce the space SBV_X of special functions with bounded X-variation in Carnot–Carathéodory spaces and study its main properties. Our main outcome is an approximation result, with respect to the BV_X topology, for SBV_X functions
Euler constants from primes in arithmetic progression
Full text linkMany Dirichlet series of number theoretic interest can be written as a
product of generating series , with ranging over all the primes in the
primitive residue class modulo , and a function well-behaved
around . In such a case the corresponding Euler constant can be expressed
in terms of the Euler constants of the series
involved and the (numerically more harmless) term . Here we
systematically study , their numerical evaluation and discuss some
examples