Search results for "methods"
showing 10 items of 4526 documents
Classical Training Methods
2006
This chapter reviews classical training methods for multilayer neural networks. These methods are widely used for classification and function modelling tasks. Nevertheless, they show a number of flaws or drawbacks that should be addressed in the development of such systems. They work by searching the minimum of an error function which defines the optimal behaviour of the neural network. Different standard problems are used to show the capabilities of these models; in particular, we have benchmarked the algorithms in a nonlinear classification problem and in three function modelling problems.
Neural network prediction in a system for optimizing simulations
2002
Neural networks have been widely used for both prediction and classification. Back-propagation is commonly used for training neural networks, although the limitations associated with this technique are well documented. Global search techniques such as simulated annealing, genetic algorithms and tabu search have also been used for this purpose. The developers of these training methods, however, have focused on accuracy rather than training speed in order to assess the merit of new proposals. While speed is not important in settings where training can be done off-line, the situation changes when the neural network must be trained and used on-line. This is the situation when a neural network i…
Estimation of Granger causality through Artificial Neural Networks: applications to physiological systems and chaotic electronic oscillators
2021
One of the most challenging problems in the study of complex dynamical systems is to find the statistical interdependencies among the system components. Granger causality (GC) represents one of the most employed approaches, based on modeling the system dynamics with a linear vector autoregressive (VAR) model and on evaluating the information flow between two processes in terms of prediction error variances. In its most advanced setting, GC analysis is performed through a state-space (SS) representation of the VAR model that allows to compute both conditional and unconditional forms of GC by solving only one regression problem. While this problem is typically solved through Ordinary Least Sq…
Crowdsourced analysis of fungal growth and branching on microfluidic platforms
2021
Fungal hyphal growth and branching are essential traits that allow fungi to spread and proliferate in many environments. This sustained growth is essential for a myriad of applications in health, agriculture, and industry. However, comparisons between different fungi are difficult in the absence of standardized metrics. Here, we used a microfluidic device featuring four different maze patterns to compare the growth velocity and branching frequency of fourteen filamentous fungi. These measurements result from the collective work of several labs in the form of a competition named the “Fungus Olympics.” The competing fungi included five ascomycete species (ten strains total), two basidiomycete…
The effectiveness of compositional animation design: Evidence from eye tracking
2017
Communication donnée le 1er septembre 2017 lors du symposium : Eye tracking as a method in learning and testing with different representations(session L 8); International audience; Learners have difficulty in decomposing conventionally designed animations to obtain raw material suitable for building high quality mental models. A composition approach to designing animations based on the Animation Processing Model was developed as a principled alternative to prevailing approaches. It provides learners with pre-decomposed material that is structured and sequenced to facilitate the relation building required for effective mental model construction. Study of a compositional animation that presen…
Measurement of the atmospheric ?µ energy spectrum from 100 GeV to 200 TeV with the ANTARES telescope
2013
Atmospheric neutrinos are produced during cascades initiated by the interaction of primary cosmic rays with air nuclei. In this paper, a measurement of the atmospheric energy spectrum in the energy range 0.1-200 TeV is presented, using data collected by the ANTARES underwater neutrino telescope from 2008 to 2011. Overall, the measured flux is similar to 25 % higher than predicted by the conventional neutrino flux, and compatible with the measurements reported in ice. The flux is compatible with a single power-law dependence with spectral index gamma (meas)=3.58 +/- 0.12. With the present statistics the contribution of prompt neutrinos cannot be established.
Astrophysics with the Laser Interferometer Space Antenna
2023
Full list of authors: Amaro-Seoane, Pau; Andrews, Jeff; Sedda, Manuel Arca; Askar, Abbas.; Baghi, Quentin; Balasov, Razvan; Bartos, Imre; Bavera, Simone S.; Bellovary, Jillian; Berry, Christopher P. L.; Berti, Emanuele; Bianchi, Stefano; Blecha, Laura; Blondin, Stephane; Bogdanovic, Tamara; Boissier, Samuel; Bonetti, Matteo; Bonoli, Silvia; Bortolas, Elisa; Breivik, Katelyn; Capelo, Pedro R.; Caramete, Laurentiu; Cattorini, Federico; Charisi, Maria; Chaty, Sylvain; Chen, Xian; Chruslinska, Martyna; Chua, Alvin J. K.; Church, Ross; Colpi, Monica; D'Orazio, Daniel; Danielski, Camilla; Davies, Melvyn B.; Dayal, Pratika; De Rosa, Alessandra; Derdzinski, Andrea; Destounis, Kyriakos; Dotti, Massi…
Absolute kinematics of radio-source components in the complete S5 polar cap sample: IV. Proper motions of the radio cores over a decade and spectral …
2016
We have carried out a high-precision astrometric analysis of two very-long-baseline-interferometry (VLBI) epochs of observation of the 13 extragalactic radio sources in the complete S5 polar cap sample. The VLBI epochs span a time baseline of ten years and enable us to achieve precisions in the proper motions of the source cores up to a few micro-arcseconds per year. The observations were performed at 14.4 GHz and 43.1 GHz, and enable us to estimate the frequency core-shifts in a subset of sources, for which the spectral-index distributions can be computed. We study the source-position stability by analysing the changes in the relative positions of fiducial source points (the jet cores) ove…
Polarization calibration techniques for the new-generation VLBI
2020
The calibration and analysis of polarization observations in Very Long Baseline Interferometry (VLBI) requires the use of specific algorithms that suffer from several limitations, closely related to assumptions in the data properties that may not hold in observations taken with new-generation VLBI equipment. Nowadays, the instantaneous bandwidth achievable with VLBI backends can be as high as several GHz, covering several radio bands simultaneously. In addition, the sensitivity of VLBI observations with state-of-the-art equipment may reach dynamic ranges of tens of thousands, both in total intensity and in polarization. In this paper, we discuss the impact of the limitations of common VLBI …
Reconstruction of inclined air showers detected with the Pierre Auger Observatory
2014
We describe the method devised to reconstruct inclined cosmic-ray air showers with zenith angles greater than $60^\circ$ detected with the surface array of the Pierre Auger Observatory. The measured signals at the ground level are fitted to muon density distributions predicted with atmospheric cascade models to obtain the relative shower size as an overall normalization parameter. The method is evaluated using simulated showers to test its performance. The energy of the cosmic rays is calibrated using a sub-sample of events reconstructed with both the fluorescence and surface array techniques. The reconstruction method described here provides the basis of complementary analyses including an…