Search

Your search keyword '"Anguita, Davide"' showing total 522 results
Start Over Author "Anguita, Davide"
522 results on '"Anguita, Davide"'

301. Digital Least Squares Support Vector Machines

Author

Anguita, Davide and Boni, Andrea

Abstract

This paper presents a very simple digital architecture that implements a Least-Squares Support Vector Machine. The simplicity of the whole system and its good behavior when used to solve classification problems hold good prospects for the application of such a kind of learning machines to build embedded systems.

Published
2003
Full Text
View/download PDF

302. Spectral Analysis of Electricity Demand Using Hilbert–Huang Transform.

Author

Luque, Joaquin, Anguita, Davide, Pérez, Francisco, and Denda, Robert

Subjects
*HILBERT-Huang transform, *ELECTRIC power consumption, *HILBERT transform, *ECONOMIC demand, *WAVELET transforms, *IMAGE compression
Abstract

The large amount of sensors in modern electrical networks poses a serious challenge in the data processing side. For many years, spectral analysis has been one of the most used approaches to extract physically meaningful information from a sea of data. Fourier Transform (FT) and Wavelet Transform (WT) are by far the most employed tools in this analysis. In this paper we explore the alternative use of Hilbert–Huang Transform (HHT) for electricity demand spectral representation. A sequence of hourly consumptions, spanning 40 months of electrical demand in Spain, has been used as dataset. First, by Empirical Mode Decomposition (EMD), the sequence has been time-represented as an ensemble of 13 Intrinsic Mode Functions (IMFs). Later on, by applying Hilbert Transform (HT) to every IMF, an HHT spectrum has been obtained. Results show smoother spectra with more defined shapes and an excellent frequency resolution. EMD also fosters a deeper analysis of abnormal electricity demand at different timescales. Additionally, EMD permits information compression, which becomes very significant for lossless sequence representation. A 35% reduction has been obtained for the electricity demand sequence. On the negative side, HHT demands more computer resources than conventional spectral analysis techniques. [ABSTRACT FROM AUTHOR]

Published
2020
Full Text
View/download PDF

303. Towards algorithms and models that we can trust: A theoretical perspective.

Author

Oneto, Luca, Ridella, Sandro, and Anguita, Davide

Subjects
*STATISTICAL learning, *ARTIFICIAL intelligence, *MACHINE learning, *TRUST, *ALGORITHMS
Abstract

In the last decade it became increasingly apparent the inability of technical metrics such as accuracy, sustainability, and non-regressiveness to well characterize the behavior of intelligent systems. In fact, they are nowadays requested to meet also ethical requirements such as explainability, fairness, robustness, and privacy increasing our trust in their use in the wild. Of course often technical and ethical metrics are in tension between each other but the final goal is to be able to develop a new generation of more responsible and trustworthy machine learning. In this paper, we focus our attention on machine learning algorithms and associated predictive models, questioning for the first time, from a theoretical perspective, if it is possible to simultaneously guarantee their performance in terms of both technical and ethical metrics towards machine learning algorithms that we can trust. In particular, we will investigate for the first time both theory and practice of deterministic and randomized algorithms and associated predictive models showing the advantages and disadvantages of the different approaches. For this purpose we will leverage the most recent advances coming from the statistical learning theory: Complexity-Based Methods, Distribution Stability, PAC-Bayes, and Differential Privacy. Results will show that it is possible to develop consistent algorithms which generate predictive models with guarantees on multiple trustworthiness metrics. • AI is nowadays requested to optimize both technical and ethical metrics. • Focus on ML algorithms and associated models (both deterministic and randomized). • We prove that it is possible to develop consistent algorithms and models. • We bound generalization in terms of both technical and ethical metrics. • We leverage the most recent advances coming from the statistical learning theory. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

307. Transition-aware human activity recognition using smartphones

Author

Reyes Ortiz, Jorge Luis, Oneto, Luca, Samà Monsonís, Albert, Ghio, Alessandro, Llanas Parra, Xavier, Anguita, Davide, Universitat Politècnica de Catalunya. Departament d'Enginyeria Electrònica, and Universitat Politècnica de Catalunya. CETpD -Centre d'Estudis Tecnològics per a l'Atenció a la Dependència i la Vida Autònoma

Subjects
Computer science, Cognitive Neuroscience, Wearable computer, 02 engineering and technology, Machine learning, computer.software_genre, Activity Recognition, Activity recognition, Artificial Intelligence, Smartphones, Support Vector Machines, Transitions, Computer Science Applications1707 Computer Vision and Pattern Recognition, Aprenentatge automàtic, 0202 electrical engineering, electronic engineering, information engineering, business.industry, Probabilistic logic, Enginyeria electrònica [Àrees temàtiques de la UPC], 020206 networking & telecommunications, Computer Science Applications, Support vector machine, Systems architecture, 020201 artificial intelligence & image processing, Artificial intelligence, business, computer
Abstract

This work presents the Transition-Aware Human Activity Recognition (TAHAR) system architecture for the recognition of physical activities using smartphones. It targets real-time classification with a collection of inertial sensors while addressing issues regarding the occurrence of transitions between activities and unknown activities to the learning algorithm. We propose two implementations of the architecture which differ in their prediction technique as they deal with transitions either by directly learning them or by considering them as unknown activities. This is accomplished by combining the probabilistic output of consecutive activity predictions of a Support Vector Machine (SVM) with a heuristic filtering approach. The architecture is validated over three case studies that involve data from people performing a broad spectrum of activities (up to 33), while carrying smartphones or wearable sensors. Results show that TAHAR outperforms state-of-the-art baseline works and reveal the main advantages of the architecture.

308. The Digital Kernel Perceptron

Author

Anguita, Davide, Boni, Andrea, Ridella, Sandro, Anguita, Davide, Boni, Andrea, and Ridella, Sandro

Abstract

In this paper, we show that a kernel-based perceptron can be efficiently implemented in digital hardware using very few components. Despite its simplicity, the experimental results on standard data sets show remarkable performance in terms of generalization error.

309. SVM Performance Assessment for the Control of Injection Moulding Processes and Plasticating Extrusion

Author

Anguita, Davide, Boni, Andrea, Tagliafico, Luca, Anguita, Davide, Boni, Andrea, and Tagliafico, Luca

Abstract

This paper presents the application of a new and promising learning algorithm based on kernel methods, i.e., support vector machines (SVMs), for the control of injection moulding processes and plasticating extrusion. In particular, the main purpose of this work is to assess the effectiveness of the method when applied to such kinds of industrial processes, characterised by a large number of variables and strictly correlated by nonlinear relationships. First, we analyse the injection process by developing a simplified model, then we identify it by using a support vector machine. The reference of the control system is tracked through the design of a control block based on the structure of the SVM.

310. The Digital Kernel Perceptron

Author

Anguita, Davide, Boni, Andrea, Ridella, Sandro, Anguita, Davide, Boni, Andrea, and Ridella, Sandro

Abstract

In this paper, we show that a kernel-based perceptron can be efficiently implemented in digital hardware using very few components. Despite its simplicity, the experimental results on standard data sets show remarkable performance in terms of generalization error.

311. The Digital Kernel Perceptron

Author

Anguita, Davide, Boni, Andrea, Ridella, Sandro, Anguita, Davide, Boni, Andrea, and Ridella, Sandro

Abstract

In this paper, we show that a kernel-based perceptron can be efficiently implemented in digital hardware using very few components. Despite its simplicity, the experimental results on standard data sets show remarkable performance in terms of generalization error.

312. SVM Performance Assessment for the Control of Injection Moulding Processes and Plasticating Extrusion

Author

Anguita, Davide, Boni, Andrea, Tagliafico, Luca, Anguita, Davide, Boni, Andrea, and Tagliafico, Luca

Abstract

This paper presents the application of a new and promising learning algorithm based on kernel methods, i.e., support vector machines (SVMs), for the control of injection moulding processes and plasticating extrusion. In particular, the main purpose of this work is to assess the effectiveness of the method when applied to such kinds of industrial processes, characterised by a large number of variables and strictly correlated by nonlinear relationships. First, we analyse the injection process by developing a simplified model, then we identify it by using a support vector machine. The reference of the control system is tracked through the design of a control block based on the structure of the SVM.

313. SVM Performance Assessment for the Control of Injection Moulding Processes and Plasticating Extrusion

Author

Anguita, Davide, Boni, Andrea, Tagliafico, Luca, Anguita, Davide, Boni, Andrea, and Tagliafico, Luca

Abstract

This paper presents the application of a new and promising learning algorithm based on kernel methods, i.e., support vector machines (SVMs), for the control of injection moulding processes and plasticating extrusion. In particular, the main purpose of this work is to assess the effectiveness of the method when applied to such kinds of industrial processes, characterised by a large number of variables and strictly correlated by nonlinear relationships. First, we analyse the injection process by developing a simplified model, then we identify it by using a support vector machine. The reference of the control system is tracked through the design of a control block based on the structure of the SVM.

314. The Digital Kernel Perceptron

Author

Anguita, Davide, Boni, Andrea, Ridella, Sandro, Anguita, Davide, Boni, Andrea, and Ridella, Sandro

Abstract

In this paper, we show that a kernel-based perceptron can be efficiently implemented in digital hardware using very few components. Despite its simplicity, the experimental results on standard data sets show remarkable performance in terms of generalization error.

315. The Digital Kernel Perceptron

Author

Anguita, Davide, Boni, Andrea, Ridella, Sandro, Anguita, Davide, Boni, Andrea, and Ridella, Sandro

Abstract

In this paper, we show that a kernel-based perceptron can be efficiently implemented in digital hardware using very few components. Despite its simplicity, the experimental results on standard data sets show remarkable performance in terms of generalization error.

316. SVM Performance Assessment for the Control of Injection Moulding Processes and Plasticating Extrusion

Author

Anguita, Davide, Boni, Andrea, Tagliafico, Luca, Anguita, Davide, Boni, Andrea, and Tagliafico, Luca

Abstract

This paper presents the application of a new and promising learning algorithm based on kernel methods, i.e., support vector machines (SVMs), for the control of injection moulding processes and plasticating extrusion. In particular, the main purpose of this work is to assess the effectiveness of the method when applied to such kinds of industrial processes, characterised by a large number of variables and strictly correlated by nonlinear relationships. First, we analyse the injection process by developing a simplified model, then we identify it by using a support vector machine. The reference of the control system is tracked through the design of a control block based on the structure of the SVM.

319. Do we really need a new theory to understand over-parameterization?

Author

Oneto, Luca, Ridella, Sandro, and Anguita, Davide

Subjects
*MACHINE learning, *ARTIFICIAL intelligence, *DEEP learning, *PUBLIC investments, *GENERALIZATION
Abstract

• Our ability to understand the mechanism behind the ML breakthroughs has slowed down. • We question the necessity for a new theoretical framework able to fill this lack. • We focus on over-parametrization a peculiar mechanism to increase generalization. • Over-parametrization can be understood using known theories and empirical evidences. This century saw an unprecedented increase of public and private investments in Artificial Intelligence (AI) and especially in (Deep) Machine Learning (ML). This led to breakthroughs in their practical ability to solve complex real-world problems impacting research and society at large. Instead, our ability to understand the fundamental mechanism behind these breakthroughs has slowed down because of their increased complexity, while in the past breakthroughs often emerged from foundational research. This questioned researchers about the necessity for a new theoretical framework able to help researchers catch up on this lag. One of the still not well understood mechanisms is the so-called over-parametrization, namely the ability of certain models to increase their generalization performance (reduce test error) when the number of parameters is above the interpolating threshold (zero training error). In this paper we will show that this phenomenon can be better understood using both known theories (surveying them in the process) and empirical evidences for both shallow and deep learning algorithms. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

320. Investigating over-parameterized randomized graph networks.

Author

Donghi, Giovanni, Pasa, Luca, Oneto, Luca, Gallicchio, Claudio, Micheli, Alessio, Anguita, Davide, Sperduti, Alessandro, and Navarin, Nicolò

Subjects
*GRAPH neural networks, *INFORMATION theory, *RANDOM graphs, *ENTROPY, *GENERALIZATION
Abstract

In this paper, we investigate neural models based on graph random features for classification tasks. First, we aim to understand when over parameterization, namely generating more features than the ones necessary to interpolate, may be beneficial for the generalization abilities of the resulting models. We employ two measures: one from the algorithmic stability framework and another one based on information theory. We provide empirical evidence from several commonly adopted graph datasets showing that the considered measures, even without considering task labels, can be effective for this purpose. Additionally, we investigate whether these measures can aid in the process of hyperparameters selection. The results of our empirical analysis show that the considered measures have good correlations with the estimated generalization performance of the models with different hyperparameter configurations. Moreover, they can be used to identify good hyperparameters, achieving results comparable to the ones obtained with a classic grid search. • Over-parameterization in randomized graph neural networks is often beneficial. • Algorithmic Stability via Gram matrix condition number bounds generalization error. • Feature entropy captures richness of graph random features. • Gram matrix condition number and feature entropy correlate with model performance. • These metrics can be used for model selection without supervision. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

323. Fairness in Machine Learning

Author

Oneto, Luca, Chiappa, Silvia, Kacprzyk, Janusz, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

324. Deep Randomized Neural Networks

Author

Gallicchio, Claudio, Scardapane, Simone, Kacprzyk, Janusz, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

325. Deep Learning for Graphs

Author

Bacciu, Davide, Micheli, Alessio, Kacprzyk, Janusz, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

326. Limitations of Shallow Networks

Author

Kůrková, Věra, Kacprzyk, Janusz, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

327. Learned Data Structures

Author

Ferragina, Paolo, Vinciguerra, Giorgio, Kacprzyk, Janusz, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

332. Traffic Sign Detection Using R-CNN

Author

Rehlaender, Philipp, Schroeer, Maik, Chadha, Gavneet, Schwung, Andreas, Angelov, Plamen, Series Editor, Kozma, Robert, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

337. Dropout for Recurrent Neural Networks

Author

Watt, Nathan, du Plessis, Mathys C., Angelov, Plamen, Series Editor, Kozma, Robert, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

343. Understanding Ancient Coin Images

Author

Cooper, Jessica, Arandjelović, Ognjen, Angelov, Plamen, Series Editor, Kozma, Robert, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

346. Asynchronous Stochastic Variational Inference

Author

Mohamad, Saad, Bouchachia, Abdelhamid, Sayed-Mouchaweh, Moamar, Angelov, Plamen, Series Editor, Kozma, Robert, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

350. Distributed SmSVM Ensemble Learning

Author

Hajewski, Jeff, Oliveira, Suely, Angelov, Plamen, Series Editor, Kozma, Robert, Series Editor, Oneto, Luca, editor, Navarin, Nicolò, editor, Sperduti, Alessandro, editor, and Anguita, Davide, editor

Published
2020
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results