Repozitorij Univerze v Novi Gorici

Iskanje po repozitoriju
A+ | A- | Pomoč | SLO | ENG

Iskalni niz: išči po
išči po
išči po
išči po
* po starem in bolonjskem študiju

Opcije:
  Ponastavi


1 - 10 / 49
Na začetekNa prejšnjo stran12345Na naslednjo stranNa konec
1.
Novel approach to fish classification : fractalysis and machine learning-based approach
Jothy Thampy, Mohanachandran Nair Sindhu Swapna, Sankaranarayana Iyer Sankararaman, 2024, objavljeni znanstveni prispevek na konferenci

Ključne besede: fractals, fish, machine learning
Objavljeno v RUNG: 15.04.2024; Ogledov: 364; Prenosov: 1
.pdf Celotno besedilo (813,33 KB)
Gradivo ima več datotek! Več...

2.
A sublexicon approach to the paradigm cell filling problem : lecture at the 5th American International Morphology Meeting, 29. 8. 2021, on-line
Guy Tabachnick, 2001, prispevek na konferenci brez natisa

Opis: How do learners figure out an inflected form of a word when they haven’t seen it before and a language allows for more than one option? In some cases, learners can make generalizations about a word’s phonological form (e.g. English verbs ending in [ɪŋ] like sting often have past tenses with [ʌŋ]). In others, as Ackerman et al. (2009) and Ackerman and Malouf (2013) show, knowing some of a word’s inflected forms often allows one to efficiently solve the Paradigm Cell Filling Problem—that is, predicting an additional form. They argue for a morphological model in which the paradigm is a fundamental unit of structure. I propose a model for how learners may use some forms of a word to predict others outside a paradigm-based formal system. In particular, I extend the sublexicon model (Gouskova et al., 2015; Becker and Gouskova, 2016), used for capturing phonological generalizations, to include dependencies between morphophonological behaviors. This can account for Hungarian possessive allomorphy, in which a noun’s choice of possessive suffix can be substantially, but not entirely, predicted both by its phonological characteristics and its membership in a certain morphological class.
Ključne besede: lexically specified allomorphy, rules of exponence, Paradigm Cell Filling Problem, sublexicons, morphological learning
Objavljeno v RUNG: 04.03.2024; Ogledov: 396; Prenosov: 2
URL Povezava na datoteko
Gradivo ima več datotek! Več...

3.
Bodies of noise at the Bell Laboratories : early automated speech recognition, contribution at the Editorial Workshop - A Special Issue on Acoustic Space, November 9-10, 2022, Frankfurt/Main
Eszter Polónyi, 2022, druga izvedena dela

Opis: This paper is about the first automated systems developed to recognize identity. While automated recognition in the twenty-first century is widely associated with images of the human face, its roots are to be found in attempts to visualize identity in other, non-figural types of trace left by human bodies, ranging as widely as shadows, astrological signs, handwriting, the prints left by palms and fingers and the acoustics of the human voice. This paper investigates one such system of recognition as it emerged from within the telecommunications industry context in the midcentury U.S. Ostensibly built to reduce human labor and cable bandwidth, Bell Labs developed three different phone devices in the 1950s to photograph, formalize and analyze the sounds of speech as they traveled through the telephony system. And while the device called “Audrey” indeed succeeded in recognizing spoken digits, it was its failure to recognize the speech contents without prior awareness of the identity of the speaker, that is to distinguish between the individuality of the speaking “medium” and their intended meaning, that arguably made the experiment a landmark in the history of machine-driven recognition. Accounting for the “noise” made by the body and the environment from which sound emanated into the device, which the lab’s technicians defined as ranging from “speech defects” to “inflection” and “background interference” proved more important than phonetic analysis in determining the intended message of given speech spectogram. Similarly to a range of experiments with noise by formalist filmmakers such as Tony Conrad, John Cage, Kurt Kren and others, it was on the principle of contingency and irreproducible uniqueness that Bell Lab technicians sought to train machine-driven intelligence.
Ključne besede: History of computer science, machine learning, Bell Labs, history of telecommunications, sound studies
Objavljeno v RUNG: 19.02.2024; Ogledov: 489; Prenosov: 6
.pdf Celotno besedilo (31,80 MB)

4.
Deep-learning-based cosmic-ray mass reconstruction using the water-Cherenkov and scintillation detectors of AugerPrime
Niklas Langner, Andrej Filipčič, Jon Paul Lundquist, Shima Ujjani Shivashankara, Samo Stanič, Serguei Vorobiov, Danilo Zavrtanik, Marko Zavrtanik, 2023, objavljeni znanstveni prispevek na konferenci

Opis: At the highest energies, cosmic rays can be detected only indirectly by the extensive air showers they create upon interaction with the Earth’s atmosphere. While high-statistics measurements of the energy and arrival directions of cosmic rays can be performed with large surface detector arrays like the Pierre Auger Observatory, the determination of the cosmic-ray mass on an event-by-event basis is challenging. Meaningful physical observables in this regard include the depth of maximum of air-shower profiles, which is related to the mean free path of the cosmic ray in the atmosphere and the shower development, as well as the number of muons that rises with the number of nucleons in a cosmic-ray particle. In this contribution, we present an approach to determine both of these observables from combined measurements of water-Cherenkov detectors and scintillation detectors, which are part of the AugerPrime upgrade of the Observatory. To characterize the time-dependent signals of the two detectors both separately as well as in correlation to each other, we apply deep learning techniques. Transformer networks employing the attention mechanism are especially well-suited for this task. We present the utilized network concepts and apply them to simulations to determine the precision of the event-by-event mass reconstruction that can be achieved by the combined measurements of the depth of shower maximum and the number of muons.
Ključne besede: Pierre Auger Observatory, ultra-high energy cosmic rays, muons, extensive air showers, surface detectors, AugerPrime, deep learning techiniques
Objavljeno v RUNG: 23.01.2024; Ogledov: 468; Prenosov: 7
.pdf Celotno besedilo (2,93 MB)
Gradivo ima več datotek! Več...

5.
Search for EeV photon-induced events at the Telescope Array
I. Kharuk, R. U. Abbasi, Y. Abe, T. Abu-Zayyad, M. Allen, Yasuhiko Arai, R. Arimura, E. Barcikowski, J. W. Belz, Douglas R. Bergman, 2023, objavljeni znanstveni prispevek na konferenci

Opis: We report on the updated results on the search for photon-like-induced events in the data, collected by Telescope Array's Surface Detectors during the last 14 years. In order to search for photon-like-induced events, we trained a neural network on Monte-Carlo simulated data to distinguish between the proton-induced and photon-induced air showers. Both reconstructed composition-sensitive parameters and raw signals registered by the Surface Detectors are used as input data for the neural network. The classification threshold was optimized to provide the strongest possible constraint on the photons' flux.
Ključne besede: Telescope Array, indirect detection, surface detection, ground array, ultra-high energy, cosmic rays, photons, neural network, machine learning
Objavljeno v RUNG: 09.10.2023; Ogledov: 847; Prenosov: 6
.pdf Celotno besedilo (543,46 KB)
Gradivo ima več datotek! Več...

6.
An improved harmony search algorithm using opposition-based learning and local search for solving the maximal covering location problem
Soumen Atta, 2023, izvirni znanstveni članek

Opis: In this article, an improved harmony search algorithm (IHSA) that utilizes opposition-based learning is presented for solving the maximal covering location problem (MCLP). The MCLP is a well-known facility location problem where a fixed number of facilities are opened at a given potential set of facility locations such that the sum of the demands of customers covered by the open facilities is maximized. Here, the performance of the harmony search algorithm (HSA) is improved by incorporating opposition-based learning that utilizes opposite, quasi-opposite and quasi-reflected numbers. Moreover, a local search heuristic is used to improve the performance of the HSA further. The proposed IHSA is employed to solve 83 real-world MCLP instances. The performance of the IHSA is compared with a Lagrangean/surrogate relaxation-based heuristic, a customized genetic algorithm with local refinement, and an improved chemical reaction optimization-based algorithm. The proposed IHSA is found to perform well in solving the MCLP instances.
Ključne besede: maximal covering location problem, harmony search algorithm, opposition-based learning, facility location problem, opposite number
Objavljeno v RUNG: 05.10.2023; Ogledov: 939; Prenosov: 7
.pdf Celotno besedilo (2,69 MB)
Gradivo ima več datotek! Več...

7.
Extraction of the Muon Signals Recorded with the Surface Detector of the Pierre Auger Observatory Using Recurrent Neural Networks
J.M. Carceller, Andrej Filipčič, Jon Paul Lundquist, Samo Stanič, Serguei Vorobiov, Danilo Zavrtanik, Marko Zavrtanik, Lukas Zehrer, 2022, objavljeni znanstveni prispevek na konferenci

Opis: We present a method based on the use of Recurrent Neural Networks to extract the muon component from the time traces registered with water-Cherenkov detector (WCD) stations of the Surface Detector of the Pierre Auger Observatory. The design of the WCDs does not allow to separate the contribution of muons to the time traces obtained from the WCDs from those of photons, electrons and positrons for all events. Separating the muon and electromagnetic components is crucial for the determination of the nature of the primary cosmic rays and properties of the hadronic interactions at ultra-high energies. We trained a neural network to extract the muon and the electromagnetic components from the WCD traces using a large set of simulated air showers, with around 450 000 simulated events. For training and evaluating the performance of the neural network, simulated events with energies between 10^18.5 eV and 10^20 eV and zenith angles below 60 degrees were used. We also study the performance of this method on experimental data of the Pierre Auger Observatory and show that our predicted muon lateral distributions agree with the parameterizations obtained by the AGASA collaboration.
Ključne besede: Pierre Auger Observatory, indirect detection, surface detection, ground array, ultra-high energy, cosmic rays, muons, machine learning, recurrent neural network
Objavljeno v RUNG: 04.10.2023; Ogledov: 814; Prenosov: 6
.pdf Celotno besedilo (1,08 MB)
Gradivo ima več datotek! Več...

8.
Telescope Array Surface Detector Energy and Arrival Direction Estimation Using Deep Learning
O. Kalashev, Jon Paul Lundquist, 2022, objavljeni znanstveni prispevek na konferenci

Opis: A novel ultra-high-energy cosmic rays energy and arrival direction reconstruction method for Telescope Array surface detector is presented. The analysis is based on a deep convolutional neural network using detector signal time series as the input and the network is trained on a large Monte-Carlo dataset. This method is compared in terms of statistical and systematic energy and arrival direction determination errors with the standard Telescope Array surface detector event reconstruction procedure.
Ključne besede: Telescope Array, indirect detection, surface detection, ground array, ultra-high energy, cosmic rays, energy, arrival directions, reconstruction, machine learning, neural network
Objavljeno v RUNG: 04.10.2023; Ogledov: 724; Prenosov: 6
.pdf Celotno besedilo (1,10 MB)
Gradivo ima več datotek! Več...

9.
Mass composition anisotropy with the Telescope Array Surface Detector data
Y. Zhezher, Jon Paul Lundquist, 2022, objavljeni znanstveni prispevek na konferenci

Opis: Mass composition anisotropy is predicted by a number of theories describing sources of ultra-high-energy cosmic rays. Event-by-event determination of a type of a primary cosmic-ray particle is impossible due to large shower-to-shower fluctuations, and the mass composition usually is obtained by averaging over some composition-sensitive observable determined independently for each extensive air shower (EAS) over a large number of events. In the present study we propose to employ the observable ξ used in the TA mass composition analysis for the mass composition anisotropy analysis. The ξ variable is determined with the use of Boosted Decision Trees (BDT) technique trained with the Monte-Carlo sets, and the ξ value is assigned for each event, where ξ=1 corresponds to an event initiated by the primary iron nuclei and ξ=−1 corresponds to a proton event. Use of ξ distributions obtained for the Monte-Carlo sets allows us to separate proton and iron candidate events from a data set with some given accuracy and study its distributions over the observed part of the sky. Results for the TA SD 11-year data set mass composition anisotropy will be presented.
Ključne besede: Telescope Array, indirect detection, surface detection, ground array, ultra-high energy, cosmic rays, composition, anisotropy, machine learning, boosted decision tree
Objavljeno v RUNG: 04.10.2023; Ogledov: 663; Prenosov: 5
.pdf Celotno besedilo (1,14 MB)
Gradivo ima več datotek! Več...

10.
Cosmic-ray mass composition with the TA SD 12-year data
Y. Zhezher, Jon Paul Lundquist, 2022, objavljeni znanstveni prispevek na konferenci

Opis: Telescope Array (TA) is the largest ultra-high-energy cosmic-ray (UHECR) observatory in the Northern Hemisphere. It is dedicated to detect extensive air showers (EAS) in hybrid mode, both by measuring the shower’s longitudinal profile with fluorescence telescopes and their particle footprint on the ground from the surface detector (SD) array. While fluorescence telescopes can measure the most composition-sensitive characteristic of EAS, the depth of the shower maximum (\xmax), they also have the drawback of small duty cycle. This work aims to study the UHECR composition based solely on the surface detector data. For this task, a set of composition-sensitive observables obtained from the SD data is used in a machine-learning method -- the Boosted Decision Trees. We will present the results of the UHECR mass composition based on the 12-year data from the TA SD using this technique, and we will discuss of the possible systematics imposed by the hadronic interaction models.
Ključne besede: Telescope Array, indirect detection, surface detection, ground array, ultra-high energy, cosmic rays, composition, machine learning, boosted decision tree
Objavljeno v RUNG: 04.10.2023; Ogledov: 648; Prenosov: 7
.pdf Celotno besedilo (763,42 KB)
Gradivo ima več datotek! Več...

Iskanje izvedeno v 0.06 sek.
Na vrh