A5006496977- Parallel Computing and Optimization Techniques
- Lattice Boltzmann Simulation Studies
- Theoretical and Computational Physics
- Material Dynamics and Properties
- Advanced Data Storage Technologies
- Rheology and Fluid Dynamics Studies
- Distributed and Parallel Computing Systems
- Fluid Dynamics and Turbulent Flows
- Advanced Malware Detection Techniques
- Stochastic processes and financial applications
- Complex Network Analysis Techniques
- Matrix Theory and Algorithms
- Complex Systems and Time Series Analysis
- User Authentication and Security Systems
- IPv6, Mobility, Handover, Networks, Security
- Wireless Networks and Protocols
- Pickering emulsions and particle stabilization
- Graph Theory and Algorithms
- Mathematical and Theoretical Epidemiology and Ecology Models
- Nanopore and Nanochannel Transport Studies
- T-cell and B-cell Immunology
- Quantum Chromodynamics and Particle Interactions
- Human Mobility and Location-Based Analysis
- Adversarial Robustness in Machine Learning
- Generative Adversarial Networks and Image Synthesis
Istituto per le Applicazioni del Calcolo Mauro Picone
2015-2024
National Research Council
2012-2024
Institute for High Performance Computing and Networking
2014-2023
Nvidia (United States)
2020
Consorzio Roma Ricerche
2019
Confederazione Nazionale dell'Artigianato e Della Piccola e Media Impresa
2015
IAC (United States)
1998-2014
National Academies of Sciences, Engineering, and Medicine
2001-2014
Sapienza University of Rome
2008
University of Catania
2007
We present a new approach to the study of immune system that combines techniques systems biology with information provided by data-driven prediction methods. To this end, we have extended an agent-based simulator response, C-ImmSim, such it represents pathogens, as well lymphocytes receptors, means their amino acid sequences and makes use bioinformatics methods for T B cell epitope prediction. This is key step simulation because determines immunogenicity. The binding epitope, which...
We investigate the security of split learning---a novel collaborative machine learning framework that enables peak performance by requiring minimal resource consumption. In present paper, we expose vulnerabilities protocol and demonstrate its inherent insecurity introducing general attack strategies targeting reconstruction clients' private training sets. More prominently, show a malicious server can actively hijack process distributed model bring it into an insecure state inference attacks...
Abstract We describe the porting of Lattice Boltzmann component MUPHY, a multi‐physics/scale simulation software, to multiple graphics processing units using Compute Unified Device Architecture. The novelty this work is development ad hoc techniques for optimizing indirect addressing that MUPHY uses efficient simulations irregular domains. Copyright © 2009 John Wiley & Sons, Ltd.
A multi-component lattice Boltzmann model recently introduced (R. Benzi et al. Phys. Rev. Lett 102, 026002 (2009)) to describe some dynamical behaviors of soft-flowing materials is theoretically analyzed. Equilibrium and transport properties are derived within the framework a continuum free-energy formulation, checked against numerical simulations. Due competition between short-range inter-species repulsion mid-range intra-species attraction, shown give rise very rich configurational...
Learning useful representations from unstructured data is one of the core challenges, as well a driving force, modern data-driven approaches. Deep learning has demonstrated broad advantages and harnessing such representations.In this paper, we introduce deep generative model representation approach for password guessing. We show that an abstract naturally offers compelling versatile properties open new directions in extensively studied, yet presently active, guessing field. These can...
A systematic study for a single-species lattice Boltzmann model with frustrated-short range attractive and mid/long-range repulsive-interactions is presented. The equilibrium analysis, carried out along the guidelines proposed by [X. Shan, Phys. Rev. E: Stat., Nonlinear, Soft Matter Phys., 2008, 77, 066702], allows us to determine surface tension density resulting disjoining pressure developing in thin film when two interfaces overlap. Numerical simulations of confined flows are then...
We discuss the Lattice Boltzmann-Particle Dynamics (LBPD) multiscale paradigm for simulation of complex states flowing matter at interface between Physics, Chemistry and Biology. In particular, we describe current large-scale LBPD simulations biopolymer translocation across cellular membranes, molecular transport in ion channels amyloid aggregation cells. also provide prospects future explorations direction organization, direct full biological organelles, all way to up physiological scales...
Abstract Quantum annealers are commercial devices that aim to solve very hard computational problems 1 , typically those involving spin glasses 2,3 . Just as in metallurgic annealing, which a ferrous metal is slowly cooled 4 quantum seek good solutions by removing the transverse magnetic field at lowest possible temperature. Removing diminishes fluctuations but forces system traverse critical point separates disordered phase (at large fields) from spin-glass small fields). A full...
Linear solvers are key components in any software platform for scientific and engineering computing. The solution of large sparse linear systems lies at the core physics-driven numerical simulations relying on partial differential equations (PDEs) often represents a significant bottleneck datadriven procedures, such as machine learning. In this paper, we present an efficient implementation preconditioned s-step Conjugate Gradient (CG) method, originally proposed by Chronopoulos Gear 1989,...
Motivation: Highly Active AntiRetroviral Therapies (HAART) can prolong life significantly to people infected by HIV since, although unable eradicate the virus, they are quite effective in maintaining control of infection. However, since HAART have several undesirable side effects, it is considered useful suspend therapy according a suitable schedule Structured Therapeutic Interruptions (STI). In present article we describe an application genetic algorithms (GA) aimed at finding optimal for...
We present a detailed analysis of the UNIX system calls and classify them according to their level threat with respect penetration. Based on these results, an effective mechanism is proposed control invocation critical, from security viewpoint, calls. The integration into existing operating systems carried out by instrumenting code in such way that execution granted only case where invoking process value arguments comply rules held access database. This method does not require changes kernel...
We present a detailed description of biopolymer translocation through nanopore in the presence solvent using an innovative multiscale methodology that treats at microscopic scale as combined with self-consistent mesoscopic for fluid dynamics. report evidence quantized current blockade depending on folding configuration and offer information role hydrodynamic correlations speeding up process.
We present the first large-scale simulation of blood flow in coronary artieries and other vessels supplying to heart muscle, with a realistic description human arterial geometry at spatial resolutions from centimeters down 10 microns (near size red cells). This multiscale resolves fluid into billion volume units, embedded bounding space 300 voxels, coupled concurrent motion million cells, which interact one another surrounding fluid. The level detail is sufficient describe phenomena...
The efficiency of lymph nodes depends on tissue structure and organization, which allow the coordination lymphocyte traffic. Despite their essential role, our understanding node specific mechanisms is still incomplete currently a topic intense research. In this paper, we present hybrid discrete/continuous model node, accounting for differences in cell velocity chemotactic response, influenced by spatial compartmentalization regulation cells migration, encounter, antigen presentation during...
Out of equilibrium relaxation processes show aging if they become slower as time passes. Aging are ubiquitous and play a fundamental role in the physics glasses spin other applications (e.g. algorithms minimizing complex cost/loss functions). The theory out dynamics mean-field glass models has achieved role, thanks to asymptotic analytic solution found by Cugliandolo Kurchan. However this is based on assumptions weak ergodicity breaking hypothesis) which have never been put under strong test...
We release a set of GPU programs for the study Quantum ($S=1/2$) Spin Glass on square lattice, with binary couplings. The library contains two main codes: MCQSG (that carries out Monte Carlo simulations using both Metropolis and Parallel Tempering algorithms, problem formulated in Trotter-Suzuki approximation), EDQSG obtains extremal eigenvalues Transfer Matrix Lanczos algorithm). has allowed us to diagonalize transfer matrices size up $2^{36}\times2^{36}$. From its side, running four NVIDIA...
Ï ÔÖÓÔÓ×Ó×ع « Ø Ú Ñ Ò ×ѸØÓ ÓÒØÖÓÐ ÒÚÓ¹