3,223 research outputs found
EVAD: encrypted vibrational anomaly detection with homomorphic encryption
No full textOne of the main concerns of cloud-based services based on machine and deep learning algorithms is the privacy of users’ data. This is particularly relevant when companies want to leverage such services because they have to outsource potentially sensible data to be processed. In this work, the problem of privacy-preserving anomaly detection on industrial vibrational data with machine learning is tackled. It consists in the detection of irregularities or deviations from expected patterns in the vibration signals generated by industrial machinery and equipment. Such anomalies can be indicative of potential equipment failures, maintenance needs, or process deviations, making their timely detection critical for ensuring the smooth operation and reliability of industrial systems. We combine this industrial need with the ability to guarantee data privacy by proposing encrypted vibrational anomaly detection (EVAD). EVAD allows the detection of anomalies on vibrational data in a privacy-preserving manner by integrating, for the first time in the literature, one-class support vector machines and homomorphic encryption, the latter being a particular kind of encryption that allows the computation of some operations directly on encrypted data. Experimental results show that, on two publicly available datasets for vibrational anomaly detection, EVAD is able to distinguish, in a privacy-preserving manner, between nominal and anomaly situations, in an effective and efficient way. To the best of our knowledge, EVAD represents the first privacy-preserving solution for the detection of anomalies in vibrational data present in the literature
Training Encrypted Neural Networks on Encrypted Data with Fully Homomorphic Encryption
No full textTraining machine and deep learning models on encrypted data is the next challenge in the field of privacy-preserving Machine and Deep Learning. The related literature in this field is very limited, since most of the solutions focus only on inference on encrypted data (leaving the training to be carried out on plain data). In this paper we introduce a multi-class and non-linear family of neural networks based on the Torus Fully Homomorphic Encryption (TFHE) scheme (named TFHE-NNs), which can be entirely trained on encrypted data. The proposed learning procedure, implementing a TFHE-compliant version of the Direct-Feedback-Alignment algorithm, is combined with a novel Cross-Validation procedure able to operate on encrypted models and encrypted accuracy. The experimental results demonstrate the feasibility of the proposed solution. The proposed models and algorithms are made available to the scientific community as a public repository
A imagem de Alessandro Baricco no Brasil
No full textDissertação (mestrado) - Universidade Federal de Santa Catarina, Centro de Comunicação e Expressão, Programa de Pós-Graduação em Estudos da Tradução, Florianópolis, 2013.Com a intenção de delinear o modo pelo qual o escritor italiano Alessandro Baricco se inseriu no sistema literário brasileiro e os caminhos percorridos pelos seus livros traduzidos, esta dissertação dá voz às experiências tradutórias de seus tradutores. A inserção de Bariccono Brasil tem seu início em 1997, através de uma proposição da Profa. Dra. Roberta Barni à editora Iluminuras da tradução de Oceano Mare. A partir daí, outras sete obras foram publicadas no Brasil, sendo três delas traduzidas por Roberta Barni e as outras quatro por quatro tradutores diferentes. De um lado, considera-se o tradutor como figura principal namediação entre culturas, e, de outro, se analisa a realidade desta figuradentro do sistema literário, sua invisibilidade, seus limites e o exercíciode sua profissão. A pesquisa conta, ainda, com críticas e resenhas referentes ao autor italiano publicadas em jornais consagrados no Brasil, considerando estas como parte constituinte da imagem de Baricco refletida em território nacional. Abstract : Intending to delineate the way the Italian writer Alessandro Baricco has been inserted in the Brazilian literary system and the paths his translated books have followed, this thesis gives voice to the translating experiences of his translators. Baricco's insertion in Brazil began in 1997, through a personal project of Dr. Roberta Barni, with her translation of Oceano Mare. Since then, seven other of his works have been published in Brazil, three of which were translated by Roberta Barni and the other four by four different translators. On the one hand,the translator is considered as the main figure in mediation betweencultures and, on the other, this figure's reality is analyzed within theliterary system: its invisibility, its limits and its professional practice. Criticisms and reviews of this Italian author published in well established Brazilian newspapers are also considered, with the understanding that they are part of Baricco's image reflected here
TIFeD: a Tiny Integer-based Federated learning algorithm with Direct feedback alignment
No full textTraining machine and deep learning models directly on extremely resource-constrained devices is the next challenge in the field of tiny machine learning. The related literature in this field is very limited, since most of the solutions focus only on on-device inference or model adaptation through online learning, leaving the training to be carried out on external Cloud services. An interesting technological perspective is to exploit Federated Learning (FL), which allows multiple devices to collaboratively train a shared model in a distributed way. However, the main drawback of state-of-the-art FL algorithms is that they are not suitable for running on tiny devices. For the first time in the literature, in this paper we introduce TIFeD, a Tiny Integer-based Federated learning algorithm with Direct Feedback Alignment (DFA) entirely implemented by using an integer-only arithmetic and being specifically designed to operate on devices with limited resources in terms of memory, computation and energy. Besides the traditional full-network operating modality, in which each device of the FL setting trains the entire neural network on its own local data, we propose an innovative single-layer TIFeD implementation, which enables each device to train only a portion of the neural network model and opens the door to a new way of distributing the learning procedure across multiple devices. The experimental results show the feasibility and effectiveness of the proposed solution. The proposed TIFeD algorithm, with its full-network and single-layer implementations, is made available to the scientific community as a public repository
Privacy-Preserving Deep Learning With Homomorphic Encryption: An Introduction
No full textPrivacy-preserving deep learning with homomorphic encryption (HE) is a novel and promising research area aimed at designing deep learning solutions that operate while guaranteeing the privacy of user data. Designing privacy-preserving deep learning solutions requires one to completely rethink and redesign deep learning models and algorithms to match the severe technological and algorithmic constraints of HE. This paper provides an introduction to this complex research area as well as a methodology for designing privacy-preserving convolutional neural networks (CNNs). This methodology was applied to the design of a privacy-preserving version of the well-known LeNet-1 CNN, which was successfully operated on two benchmark datasets for image classification. Furthermore, this paper details and comments on the research challenges and software resources available for privacy-preserving deep learning with HE
To Personalize or Not To Personalize? Soft Personalization and the Ethics of ML for Health
No full textPersonalization is among the most promising outcomes of using Machine Learning models that can be trained on data representing a specific individual. Personalization is particularly promising in areas such as health and medicine, as several crucial aspects and determinants of health are individual. Yet additional ethical issues arise with increasingly personalized models, including privacy, acceptability, reliability and trade-offs. In this paper we discuss and propose ML models for health that can be personalized on individual users, while guaranteeing both their privacy and quality from an ethical and epistemic (knowledge-related) point of view. To achieve these goals, we argue that we need to control the learning and evolution of personalized models. We propose soft personalization as an ethicallyinformed framework to limit personalization and respect epistemic and ethical values that are specific for the health context, including representativity, quality, non-maleficence, beneficence, privacy. Based on an interdisciplinary approach combining the philosophical and computer science scholarship of our group, soft personalization is a way of developing different models that can be selected depending on their quality and safety. We characterize the approach theoretically and technically and make it concrete with a case study of glucose monitoring and anomaly detection through privacy-preserving ML. Our framework shows that, even when individual issues such as privacy can be mitigated, tradeoffs with other values remain and choices are necessary as to which values should be prioritized
New echocardiographic screening tool for left ventricular tract obstruction risk assessment in TMVR
No full textBackground: Transcatheter mitral valve replacement (TMVR) is an alternative to conventional surgery to treat severe mitral disease but its use is limited by the risk of left ventricular outflow tract obstruction (LVOTO). Screening depends on ECG-gated computed tomography (CT) that is not widely available and requires contrast. We developed and validated a transthoracic echocardiographic (TTE) method to assess the risk of LVOTO after TMVR with the Tendyne System. Methods: We measured the LVOT longitudinal area on preoperative TTE dataset of patients screened for TMVR. The LVOT was measured as the box-area included by the aortic valve annulus, the anterior mitral leaflet (AML), the c-septum distance line, and the respective length of the AML on the interventricular septum. We analyzed the correlation between the TTE LVOT-box and the CT-measured neoLVOT area. Prediction performance for eligible patients was tested with ROC curves. Results: Thirty-nine patients were screened, out of 14 patients (36%) not eligible for TMVR, 8 had risk of LVOTO. We found a linear correlation between the TTE LVOT-box and the CT-measured Neo-LVOT (r = 0.6, p = 0.002). ROC curve showed that the method is specific and sensitive and the cut-off value of the measure LVOT-box is 350 mm2. Conclusions: The proposed method is reliable to evaluate the risk of LVOTO after TMR with the Tendyne System. It is quick and easy and can be used as a first-line assessment in the outpatient clinic. Patients with LVOT-box <350 mm2 should not be further screened with ECG-gated cardiac CT
La maturità di Alessandro Fei del Barbiere, in bilico tra Maniera e Riforma
No full textThis article studies the mature career of the Florentine painter Alessandro Fei del Barbiere (1537-1592), beginning with the rediscovery of the 'Ascension' altarpiece formerly in the Albizi Chapel in the destroyed church of San Pier Maggiore, Florence. Studying this painting and others recorded in 1584 by the biographer Raffaello Borghini, such as the two altarpieces for Santa Maria delle Grazie and the Madonna dell'Umiltà in Pistoia, the author reconstructs a body of works showing how in the 1580s Fei gradually went beyond the archaic style of his apprenticeship - he had been trained by Ridolfo del Ghirlandaio and Pierfrancesco Foschi, but was also marked by the Maniera of Vasari - evolving towards naturalism in both mimesis and pictorial handling. In Florence, his development partly parallels that of Santi di Tito and his circle, but Fei was also influenced by a probable sojourn during the early part of that decade in Rome, where he could have been inspired by Girolamo Muziano and the painters working for Pope Gregory XIII. Among other proposals, the author suggests that the artist was responsible for decorating the chancel of Fiesole Cathedral (c. 1584-1589), which consisted of an altarpiece, only rarely discussed by scholars, and a cycle of frescoes hitherto attributed to Nicodemo Ferrucci
- …
