A5102864628- Rough Sets and Fuzzy Logic
- Advanced Graph Neural Networks
- Complex Network Analysis Techniques
- Bioinformatics and Genomic Networks
- Computational Drug Discovery Methods
- Graph Theory and Algorithms
- Data Mining Algorithms and Applications
- Parkinson's Disease Mechanisms and Treatments
- Network Security and Intrusion Detection
- IoT Networks and Protocols
- Anomaly Detection Techniques and Applications
- Advanced Clustering Algorithms Research
- Algorithms and Data Compression
- Protein Structure and Dynamics
- Network Packet Processing and Optimization
- Privacy-Preserving Technologies in Data
- Machine Learning in Bioinformatics
- IoT and Edge/Fog Computing
- Biomedical Text Mining and Ontologies
- Bluetooth and Wireless Communication Technologies
- Human Rights and Immigration
- Topological and Geometric Data Analysis
- Misinformation and Its Impacts
- Neural Networks and Applications
- Topic Modeling
Libera Università Internazionale degli Studi Sociali Guido Carli
2021-2025
Institute of Cognitive Sciences and Technologies
2021-2022
Sapienza University of Rome
2017-2021
National Research Council
2021
Objective To apply a machine learning analysis to clinical and presynaptic dopaminergic imaging data of patients with rapid eye movement (REM) sleep behavior disorder (RBD) predict the development Parkinson disease (PD) dementia Lewy bodies (DLB). Method s In this multicenter study International RBD group, 173 (mean age 70.5 ± 6.3 years, 70.5% males) polysomnography‐confirmed who eventually phenoconverted overt alpha‐synucleinopathy (RBD due synucleinopathy) were enrolled, underwent baseline...
The year 2020 opened with a dramatic epidemic caused by new species of coronavirus that soon has been declared pandemic the WHO due to high number deaths and critical mass worldwide hospitalized patients, order millions. COVID-19 forced governments hundreds countries apply several heavy restrictions in citizens' socio-economic life. Italy was one most affected long-term restrictions, impacting tissue. During this lockdown period, people got informed mostly on Online Social Media, where...
The introduction of Transformer architectures – with the self-attention mechanism in automatic Natural Language Generation (NLG) is a breakthrough solving general task-oriented problems, such as simple production long text excerpts that resemble ones written by humans. While performance GPT-X there for all to see, many efforts are underway penetrate secrets these black-boxes terms intelligent information processing whose output statistical distributions natural language. In this work,...
One notable paradigm shift in Natural Language Processing has been the introduction of Transformers, revolutionizing language modeling as Convolutional Neural Networks did for Computer Vision. The power along with many other innovative features, also lies integration word embedding techniques, traditionally used to represent words a text and build classification systems directly. This study delves into comparison representation techniques classifying users who generate medical topic posts on...
In this paper, we propose a novel implementation for solving the large-scale k-medoids clustering problem.Conversely to most famous k-means, suffers from computationally intensive phase medoids evaluation, whose complexity is quadratic in space and time; thus task large datasets and, specifically, clusters might be unfeasible.In order overcome problem, two alternatives update, one exact method approximate method: former based on solving, distributed fashion, medoid update problem; latter...
This paper investigates a novel graph embedding procedure based on simplicial complexes. Inherited from algebraic topology, complexes are collections of increasing-order simplices (e.g., points, lines, triangles, tetrahedrons) which can be interpreted as possibly meaningful substructures (i.e., information granules) the top an space built by means symbolic histograms. In space, any Euclidean pattern recognition system used, equipped with feature selection capabilities in order to select most...
Large scale deployments of Internet Things (IoT) networks are becoming reality. From a technology perspective, lot information related to device parameters, channel states, network and application data stored in databases can be used for an extensive analysis improve the functionality IoT systems terms performance user services. LoRaWAN (Long Range Wide Area Network) is one emerging technologies, with simple protocol based on LoRa modulation. In this work, we discuss how machine learning...
MRI studies reported that ALS patients with bulbar and spinal onset showed focal cortical changes in corresponding regions of the motor homunculus. We evaluated capability brain 2-[18F]FDG-PET to disclose metabolic features characterizing pure or impairment.We classified as (PB) a normal score items ALSFRS-R, (PS) at time PET. Forty healthy controls (HC) were enrolled. compared PB PS, each patient group HC. Metabolic clusters showing statistically significant difference between PS tested...
In this paper we discuss techniques for potential speedups in k-medoids clustering. Specifically, address the advantages of pre-caching pairwise distance matrix, heart clustering algorithm, not only order to speedup execution algorithm itself, but also evaluation well-known Silhouette Index and DaviesBouldin clusters' validation. A major disadvantage such is that it might be suitable large datasets. To end, a further contribution consists proposing parallel distributed implementations both...
Topological Data Analysis is a novel approach, useful whenever data can be described by topological structures such as graphs. The aim of this paper to investigate whether tool used in order define set descriptors for pattern recognition and machine learning tasks. Specifically, we consider supervised problem with the final goal predicting proteins' physiological function starting from their respective residue contact network. Indeed, folded proteins effectively graphs, making them...
A novel energy management system (EMS) synthesis procedure based on adaptive neurofuzzy inference systems (ANFISs) by hyperplane clustering is investigated in this paper. In particular, since it known that input-output samples space does not consider clusters' separability the input space, a Min-Max classifier applied to properly cut and update those hyperplanes defined scattered clusters order refine ANFIS membership functions. Furthermore, three different techniques have been compared for...
Graphs are data structures able to efficiently describe real-world systems and, as such, have been extensively used in recent years by many branches of science, including machine learning engineering. However, the design efficient graph-based pattern recognition is bottlenecked intrinsic problem how properly match two graphs. In this paper, we investigate a granular computing approach for general purpose classification system. The overall framework relies on extraction meaningful pivotal...
Dissimilarity spaces, along with feature reduction/ selection techniques, are among the mainstream approaches when dealing pattern recognition problems in structured (and possibly non-metric) domains. In this work, we aim at investigating dissimilarity space representations a biology-related application, namely protein function classification, as proteins seminal example of data given their primary and tertiary structures. Specifically, propose two different analyses relying on both complete...
We investigated sex-related brain metabolic differences in Amyotrophic Lateral Sclerosis (ALS) and healthy controls (HC).