A5011672584- Theoretical and Computational Physics
- Neural Networks and Applications
- Complex Network Analysis Techniques
- Complex Systems and Time Series Analysis
- Neural dynamics and brain function
- Gene Regulatory Network Analysis
- Statistical Mechanics and Entropy
- Opinion Dynamics and Social Influence
- Quantum many-body systems
- Model Reduction and Neural Networks
- Advanced Thermodynamics and Statistical Mechanics
- Topological and Geometric Data Analysis
- Artificial Immune Systems Applications
- Generative Adversarial Networks and Image Synthesis
- Stochastic processes and statistical mechanics
- 3D Printing in Biomedical Research
- Advanced Memory and Neural Computing
- T-cell and B-cell Immunology
- Protein Structure and Dynamics
- Material Dynamics and Properties
- Mathematical and Theoretical Epidemiology and Ecology Models
- Evolution and Genetic Dynamics
- Neural Networks and Reservoir Computing
- Quantum chaos and dynamical systems
- Cell Image Analysis Techniques
Sapienza University of Rome
2008-2024
Istituto Nazionale di Fisica Nucleare
2007-2024
University of Salento
2013-2024
Istituto Nazionale di Fisica Nucleare, Sezione di Lecce
2018-2024
Istituto Nazionale di Alta Matematica Francesco Severi
2011-2023
Central Maine Community College
2023
Institute of Genetics and Biophysics
1997-2022
National Research Council
2017-2022
University of California, Los Angeles
2022
Istituto Nazionale di Fisica Nucleare, Sezione di Roma I
2012-2019
We introduce a bipartite, diluted and frustrated, network as sparse restricted Boltzman machine we show its thermodynamical equivalence to an associative working memory able retrieve multiple patterns in parallel without falling into spurious states typical of classical neural networks. focus on systems processing finite (up logarithmic growth the volume) amount patterns, mirroring low-level storage standard Amit-Gutfreund-Sompolinsky theory. Results obtained trough statistical mechanics,...
Aim of this paper is to give an extensive treatment bipartite mean field spin systems, ordered and disordered: at first, ferromagnets are investigated, achieving explicit expression for the free energy trough a new minimax variational principle.Furthermore via Hamilton-Jacobi technique same structure obtained together with existence its thermodynamic limit principle connected standard max one.The investigated spin-glasses: By Borel-Cantelli lemma control high temperature regime obtained,...
In this paper we discuss the applicability of numerical descriptors and statistical physics concepts to characterize complex biological systems observed at microscopic level through organ on chip approach. To end, employ data collected a microfluidic platform in which leukocytes can move suitably built channels toward their target. Leukocyte behavior is recorded by standard time lapse imaging. particular, analyze three groups human peripheral blood mononuclear cells (PBMC): heterozygous...
Scope of the present work is to infer migratory ability leukocytes by stochastic processes in order distinguish spontaneous organization immune cells against an insult (namely cancer). For this purpose, spleen from immunodeficient mice, selectively lacking transcription factor IRF-8 (IRF-8 knockout; KO), or immunocompetent animals (wild-type; WT), were allowed interact, alternatively, with murine B16.F10 melanoma ad hoc microfluidic environment developed on a LabOnChip technology. In...
This paper studies the connection between Hopfield networks and restricted Boltzmann machines, two common tools in developing area of machine learning. The work focuses on behavior models whose variables are either discrete binary or take a range continuous values. authors find large degree robustness retrieval capabilities models, especially when number stored patterns is small.
Oral-facial-digital type I (OFDI) syndrome is a male-lethal X-linked dominant developmental disorder belonging to the heterogeneous group of oral-facial-digital syndromes (OFDS). OFDI characterized by malformations face, oral cavity, and digits. Central nervous system (CNS) abnormalities cystic kidney disease can also be part this condition. This rare genetic due mutations in OFD1 gene that encodes centrosome/basal body protein necessary for primary cilium assembly left-right axis...
We adapt belief-propagation techniques to study the equilibrium behavior of a bipartite spin glass, with interactions between two sets N and P=αN spins each having an arbitrary degree, i.e., number interaction partners in opposite set. An equivalent view is then system neurons storing P diluted patterns via Hebbian learning, high storage regime. Our method allows analysis parallel pattern processing on broad class graphs, including those asymmetry heterogeneous dilution; previous replica...
Pattern-diluted associative networks were introduced recently as models for the immune system, with nodes representing T-lymphocytes and stored patterns signalling protocols between T-and B-lymphocytes.It was shown earlier that in regime of extreme pattern dilution, a system N T Tlymphocytes can manage numberHere we study this model extensive load B = αN , also high degree agreement immunological findings.We use graph theory statistical mechanical analysis based on replica methods to show...
We study generalized restricted Boltzmann machines with generic priors for units and weights, interpolating between Boolean Gaussian variables. present a complete analysis of the replica symmetric phase diagram these systems, which can be regarded as Hopfield models. underline role retrieval both inference learning processes we show that is robust large class weight unit priors, beyond standard scenario. Furthermore, how paramagnetic boundary directly related to optimal size training set...
We consider statistical-mechanics models for spin systems built on hierarchical structures, which provide a simple example of non-mean-field framework. show that the coupling decay with distance can give rise to peculiar features and phase diagrams much richer than their mean-field counterpart. In particular, we Dyson model, mimicking ferromagnetism in lattices, prove existence number metastabilities, beyond ordered state, become stable thermodynamic limit. Such feature is retained when...
Abstract In this paper we develop statistical algorithms to infer possible cardiac pathologies, based on data collected from 24 h Holter recording over a sample of 2829 labelled patients; labels highlight whether patient is suffering pathologies. the first part work analyze statistically heart-beat series associated each and them out get coarse-grained description heart variability in terms 49 markers well established reference community. These are then used as inputs for multi-layer...
Abstract In neural network's literature, Hebbian learning traditionally refers to the procedure by which Hopfield model and its generalizations store archetypes ( i.e. , definite patterns that are experienced just once form synaptic matrix). However, term in machine ability of extract features from supplied dataset e.g. made blurred examples these archetypes), order make own representation unavailable archetypes. Here, given a sample examples, we define supervised protocol based on Hebb's...
The aim of this work is to implement a statistical mechanics theory social interaction, generalizing econometric discrete choice models. A class simple mean-field models introduced and discussed both from the theoretical phenomenological point view. We propose parameter evaluation procedure test it by fitting model against three families data coming different cases: estimated interaction parameters are found have similar positive values, giving quantitative confirmation peer imitation...
The laminin alpha 5 gene (LAMA5) plays a master role in the maintenance and function of extracellular matrix (ECM) mammalian tissues, which is critical developmental patterning, stem cell niches, cancer genetic diseases. Its mutations have never been reported human disease so far. aim this study was to associate first mutation LAMA5 novel multisystem syndrome.A detailed characterisation three-generation family, including clinical, biochemical, instrumental morphological analysis, together...