Search

Your search keyword '"Golik P"' showing total 552 results
Start Over Author "Golik P"
552 results on '"Golik P"'

1. Hybrid Ensemble-Based Travel Mode Prediction

Author

Golik, Paweł, Grzenda, Maciej, and Sienkiewicz, Elżbieta

Subjects
Computer Science - Machine Learning
Abstract

Travel mode choice (TMC) prediction, which can be formulated as a classification task, helps in understanding what makes citizens choose different modes of transport for individual trips. This is also a major step towards fostering sustainable transportation. As behaviour may evolve over time, we also face the question of detecting concept drift in the data. This necessitates using appropriate methods to address potential concept drift. In particular, it is necessary to decide whether batch or stream mining methods should be used to develop periodically updated TMC models. To address the challenge of the development of TMC models, we propose the novel Incremental Ensemble of Batch and Stream Models (IEBSM) method aimed at adapting travel mode choice classifiers to concept drift possibly occurring in the data. It relies on the combination of drift detectors with batch learning and stream mining models. We compare it against batch and incremental learners, including methods relying on active drift detection. Experiments with varied travel mode data sets representing both city and country levels show that the IEBSM method both detects drift in travel mode data and successfully adapts the models to evolving travel mode choice data. The method has a higher rank than batch and stream learners., Comment: This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this contribution is published in Advances in Intelligent Data Analysis XXII. IDA 2024. Lecture Notes in Computer Science, vol 14641. Springer, and is available online at Cham https://doi.org/10.1007/978-3-031-58547-0_16 The preprint includes 12+22 pages, 1+1 figures

Published
2024
Full Text
View/download PDF

6. Theory of classical electrodynamics with topologically quantized singularities as electric charges

Author

Golik, Bruno, Jukić, Dario, and Buljan, Hrvoje

Subjects
Condensed Matter - Mesoscale and Nanoscale Physics, High Energy Physics - Theory
Abstract

We formulate a theory of classical electrodynamics where the only admissible electric charges are topological singularities in the electromagnetic field, and charge quantization is accounted by the Chern theorem, such that Dirac magnetic monopoles are not needed. The theory allows positive and negative charges of equal magnitude, where the sign of the charge corresponds to the chirality of the topological singularity. Given the trajectory ${\bf w}(t)$ of the singularity, one can calculate electric and magnetic fields identical to those produced by Maxwell's equations for a moving point charge, apart from a multiplicative constant factor related to electron charge and vacuum permittivity. The theory is based on the relativistic Weyl equation in frequency-wavevector space, with eigenstates comprising the position, velocity, and acceleration of the singularity, and eigenvalues defining the retarded position of the charge. From the eigenstates we calculate the Berry connection and the Berry curvatures, and identify the curvatures as electric and magnetic fields., Comment: are welcome

Published
2023
Full Text
View/download PDF

10. Accented Speech Recognition: Benchmarking, Pre-training, and Diverse Data

Author

Aksënova, Alëna, Chen, Zhehuai, Chiu, Chung-Cheng, van Esch, Daan, Golik, Pavel, Han, Wei, King, Levi, Ramabhadran, Bhuvana, Rosenberg, Andrew, Schwartz, Suzan, and Wang, Gary

Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing, Computer Science - Sound
Abstract

Building inclusive speech recognition systems is a crucial step towards developing technologies that speakers of all language varieties can use. Therefore, ASR systems must work for everybody independently of the way they speak. To accomplish this goal, there should be available data sets representing language varieties, and also an understanding of model configuration that is the most helpful in achieving robust understanding of all types of speech. However, there are not enough data sets for accented speech, and for the ones that are already available, more training approaches need to be explored to improve the quality of accented speech recognition. In this paper, we discuss recent progress towards developing more inclusive ASR systems, namely, the importance of building new data sets representing linguistic diversity, and exploring novel training approaches to improve performance for all users. We address recent directions within benchmarking ASR systems for accented speech, measure the effects of wav2vec 2.0 pre-training on accented speech recognition, and highlight corpora relevant for diverse ASR evaluations., Comment: 5 pages, 3 tables

Published
2022

12. Optical biosensor based on a photonic crystal with a defective layer designed to determine the concentration of SARS-CoV-2 in water

Author

Efimov, I. M., Vanyushkin, N. A., Gevorgyan, A. H., and Golik, S. S.

Subjects
Physics - Instrumentation and Detectors, Physics - Optics
Abstract

We propose a new optical biosensor based on a photonic crystal with a defect layer, which can determine the SARS-CoV-2 concentration in water by the defect mode shift. Two models of the dependence of the refractive index of the defect layer on the concentration of the pathogen in water were considered. The optimal parameters for the photonic crystal in our device and the optimal thickness of the defect layer were determined. It was also demonstrated that in the presence of absorption in the investigated structure it is much more advantageous to work in the reflection mode compared to the transmission mode. Finally, the wavelength dependence of the defect mode on the SARS-CoV-2 concentration was obtained and the sensitivity of the sensor was determined., Comment: 15 pages, 13 figures

Published
2022
Full Text
View/download PDF

13. Scattering of a plane wave by an inhomogeneous 1D dielectric layer with gradient refractive index

Author

Vanyushkin, N. A., Gevorgyan, A. H., and Golik, S. S.

Subjects
Physics - Optics
Abstract

We propose a new method for calculating reflection and transmission coefficients for an arbitrarily polarized electromagnetic plane wave incident on a one-dimensional dielectric medium of finite thickness and with dielectric permittivity being an arbitrary continuous function of the coordinate. We have shown that the problem of plane wave scattering by an inhomogeneous layer is reduced to a system of first order differential equations that contain the derivative of the refractive index or dielectric permittivity of the layer, which can be used, for example, when searching for an analytical solution. This method also makes it easy to obtain the distribution of the field strength within the layer. The reflection spectra and field distribution obtained using this method were compared with the analytical solution based on Mathieu functions., Comment: 14 pages, 11 figures

Published
2022
Full Text
View/download PDF

14. Including the Dark Side of Entrepreneurship in the Entrepreneurship Education

Author

Ziemianski, Pawel and Golik, Jakub

Abstract

Pursuing an entrepreneurial career is often rewarding in terms of both economic and psychological outcomes. However, becoming an entrepreneur also has its darker side that affects professional and personal life. Meanwhile, the positivity bias is prevalent in entrepreneurial education and research. It is recognized as emphasizing the advantages of becoming an entrepreneur and giving considerably less attention to potential downsides. Based on the theoretical model of met expectations, it is proposed that building an accurate and balanced image of the entrepreneurial career is crucial to help students prepare to pursue it successfully. Using data from SEAS (Survey on Entrepreneurial Attitudes of Students) Project, authors quantitatively test the perception of the severity of negative aspects of entrepreneurship among 513 business students from northern Poland. Further, the results of 16 semi-structured qualitative interviews conducted with mature and experienced entrepreneurs from the same region are presented. They are focused on the entrepreneurs' perspective on the experienced dark sides and reveal employed coping strategies. A call is made to include these findings in designing university entrepreneurship programs by eliciting the awareness of the existence of the dark sides and indicating the means of their attenuation.

Published
2020

16. Building Large Lexicalized Ontologies from Text: a Use Case in Automatic Indexing of Biotechnology Patents

Author

Nédellec, Claire, Golik, Wiktoria, Aubin, Sophie, and Bossy, Robert

Subjects
Computer Science - Artificial Intelligence, Computer Science - Computation and Language, Computer Science - Information Retrieval
Abstract

This paper presents a tool, TyDI, and methods experimented in the building of a termino-ontology, i.e. a lexicalized ontology aimed at fine-grained indexation for semantic search applications. TyDI provides facilities for knowledge engineers and domain experts to efficiently collaborate to validate, organize and conceptualize corpus extracted terms. A use case on biotechnology patent search demonstrates TyDI's potential.

Published
2020
Full Text
View/download PDF

17. Neural Simultaneous Speech Translation Using Alignment-Based Chunking

Author

Wilken, Patrick, Alkhouli, Tamer, Matusov, Evgeny, and Golik, Pavel

Subjects
Computer Science - Computation and Language
Abstract

In simultaneous machine translation, the objective is to determine when to produce a partial translation given a continuous stream of source words, with a trade-off between latency and quality. We propose a neural machine translation (NMT) model that makes dynamic decisions when to continue feeding on input or generate output words. The model is composed of two main components: one to dynamically decide on ending a source chunk, and another that translates the consumed chunk. We train the components jointly and in a manner consistent with the inference conditions. To generate chunked training data, we propose a method that utilizes word alignment while also preserving enough context. We compare models with bidirectional and unidirectional encoders of different depths, both on real speech and text input. Our results on the IWSLT 2020 English-to-German task outperform a wait-k baseline by 2.6 to 3.7% BLEU absolute., Comment: IWSLT 2020

Published
2020

22. Cumulative Adaptation for BLSTM Acoustic Models

Author

Kitza, Markus, Golik, Pavel, Schlüter, Ralf, and Ney, Hermann

Subjects
Computer Science - Computation and Language, Statistics - Machine Learning
Abstract

This paper addresses the robust speech recognition problem as an adaptation task. Specifically, we investigate the cumulative application of adaptation methods. A bidirectional Long Short-Term Memory (BLSTM) based neural network, capable of learning temporal relationships and translation invariant representations, is used for robust acoustic modelling. Further, i-vectors were used as an input to the neural network to perform instantaneous speaker and environment adaptation, providing 8\% relative improvement in word error rate on the NIST Hub5 2000 evaluation test set. By enhancing the first-pass i-vector based adaptation with a second-pass adaptation using speaker and environment dependent transformations within the network, a further relative improvement of 5\% in word error rate was achieved. We have reevaluated the features used to estimate i-vectors and their normalization to achieve the best performance in a modern large scale automatic speech recognition system., Comment: Submitted to Interspeech 2019

Published
2019

28. Mind the Gap: Line Managers as Effective or Non-Effective Talent Spotters

Author

Blanco, María Rita and Golik, Mariela N.

Abstract

Due to the centrality of line managers' role in the talent identification process, this qualitative study aims to determine what effective and non-effective talent spotters do differently from one another in relation with an 'ideal' talent identification model offered by the literature. A double-blind design was applied to the sample designated by the Corporate Talent Management Director of a Latin American multinational firm. 20 senior and middle managers, some effective and some others non-effective, who work in different countries, participated in the study. Findings show that effective talent spotters perform three key stages of the talent identification model and they do it extremely well: differentiation of potential from performance, matching process and design of developmental challenges. The clear differentiation of potential from performance constitutes the crucial stage since a fuzzy differentiation taints the following stages. Not only is it necessary to perform these key stages but they should be executed in line with the principles of the talent management literature. Recommendations for practice, including specific guidelines for Talent Managers are proposed.

Published
2021
Full Text
View/download PDF

33. A comprehensive study of batch construction strategies for recurrent neural networks in MXNet

Author

Doetsch, Patrick, Golik, Pavel, and Ney, Hermann

Subjects
Computer Science - Learning, Statistics - Machine Learning
Abstract

In this work we compare different batch construction methods for mini-batch training of recurrent neural networks. While popular implementations like TensorFlow and MXNet suggest a bucketing approach to improve the parallelization capabilities of the recurrent training process, we propose a simple ordering strategy that arranges the training sequences in a stochastic alternatingly sorted way. We compare our method to sequence bucketing as well as various other batch construction strategies on the CHiME-4 noisy speech recognition corpus. The experiments show that our alternated sorting approach is able to compete both in training time and recognition performance while being conceptually simpler to implement.

Published
2017

48. To which world regions does the valence–dominance model of social perception apply?

Author

Jones, Benedict C., DeBruine, Lisa M., Flake, Jessica K., Liuzza, Marco Tullio, Antfolk, Jan, Arinze, Nwadiogo C., Ndukaihe, Izuchukwu L. G., Bloxsom, Nicholas G., Lewis, Savannah C., Foroni, Francesco, Willis, Megan L., Cubillas, Carmelo P., Vadillo, Miguel A., Turiegano, Enrique, Gilead, Michael, Simchon, Almog, Saribay, S. Adil, Owsley, Nicholas C., Jang, Chaning, Mburu, Georgina, Calvillo, Dustin P., Wlodarczyk, Anna, Qi, Yue, Ariyabuddhiphongs, Kris, Jarukasemthawee, Somboon, Manley, Harry, Suavansri, Panita, Taephant, Nattasuda, Stolier, Ryan M., Evans, Thomas R., Bonick, Judson, Lindemans, Jan W., Ashworth, Logan F., Hahn, Amanda C., Chevallier, Coralie, Kapucu, Aycan, Karaaslan, Aslan, Leongómez, Juan David, Sánchez, Oscar R., Valderrama, Eugenio, Vásquez-Amézquita, Milena, Hajdu, Nandor, Aczel, Balazs, Szecsi, Peter, Andreychik, Michael, Musser, Erica D., Batres, Carlota, Hu, Chuan-Peng, Liu, Qing-Lan, Legate, Nicole, Vaughn, Leigh Ann, Barzykowski, Krystian, Golik, Karolina, Schmid, Irina, Stieger, Stefan, Artner, Richard, Mues, Chiel, Vanpaemel, Wolf, Jiang, Zhongqing, Wu, Qi, Marcu, Gabriela M., Stephen, Ian D., Lu, Jackson G., Philipp, Michael C., Arnal, Jack D., Hehman, Eric, Xie, Sally Y., Chopik, William J., Seehuus, Martin, Azouaghe, Soufian, Belhaj, Abdelkarim, Elouafa, Jamal, Wilson, John P., Kruse, Elliott, Papadatou-Pastou, Marietta, De La Rosa-Gómez, Anabel, Barba-Sánchez, Alan E., González-Santoyo, Isaac, Hsu, Tsuyueh, Kung, Chun-Chia, Wang, Hsiao-Hsin, Freeman, Jonathan B., Oh, Dong Won, Schei, Vidar, Sverdrup, Therese E., Levitan, Carmel A., Cook, Corey L., Chandel, Priyanka, Kujur, Pratibha, Parganiha, Arti, Parveen, Noorshama, Pati, Atanu Kumar, Pradhan, Sraddha, Singh, Margaret M., Pande, Babita, Bavolar, Jozef, Kačmár, Pavol, Zakharov, Ilya, Álvarez-Solas, Sara, Baskin, Ernest, Thirkettle, Martin, Schmidt, Kathleen, Christopherson, Cody D., Leonis, Trinity, Suchow, Jordan W., Olofsson, Jonas K., Jernsäther, Teodor, Lee, Ai-Suan, Beaudry, Jennifer L., Gogan, Taylor D., Oldmeadow, Julian A., Balas, Benjamin, Stevens, Laura M., Colloff, Melissa F., Flowe, Heather D., Gülgöz, Sami, Brandt, Mark J., Hoyer, Karlijn, Jaeger, Bastian, Ren, Dongning, Sleegers, Willem W. A., Wissink, Joeri, Kaminski, Gwenaël, Floerke, Victoria A., Urry, Heather L., Chen, Sau-Chin, Pfuhl, Gerit, Vally, Zahir, Basnight-Brown, Dana M., Jzerman, Hans I., Sarda, Elisa, Neyroud, Lison, Badidi, Touhami, Van der Linden, Nicolas, Tan, Chrystalle B. Y., Kovic, Vanja, Sampaio, Waldir, Ferreira, Paulo, Santos, Diana, Burin, Debora I., Gardiner, Gwendolyn, Protzko, John, Schild, Christoph, Ścigała, Karolina A., Zettler, Ingo, O’Mara Kunz, Erin M., Storage, Daniel, Wagemans, Fieke M. A., Saunders, Blair, Sirota, Miroslav, Sloane, Guyan V., Lima, Tiago J. S., Uittenhove, Kim, Vergauwe, Evie, Jaworska, Katarzyna, Stern, Julia, Ask, Karl, van Zyl, Casper J. J., Körner, Anita, Weissgerber, Sophia C., Boudesseul, Jordane, Ruiz-Dodobara, Fernando, Ritchie, Kay L., Michalak, Nicholas M., Blake, Khandis R., White, David, Gordon-Finlayson, Alasdair R., Anne, Michele, Janssen, Steve M. J., Lee, Kean Mun, Nielsen, Tonje K., Tamnes, Christian K., Zickfeld, Janis H., Rosa, Anna Dalla, Vianello, Michelangelo, Kocsor, Ferenc, Kozma, Luca, Putz, Ádám, Tressoldi, Patrizio, Irrazabal, Natalia, Chatard, Armand, Lins, Samuel, Pinto, Isabel R., Lutz, Johannes, Adamkovic, Matus, Babincak, Peter, Baník, Gabriel, Ropovik, Ivan, Coetzee, Vinet, Dixson, Barnaby J. W., Ribeiro, Gianni, Peters, Kim, Steffens, Niklas K., Tan, Kok Wei, Thorstenson, Christopher A., Fernandez, Ana Maria, Hsu, Rafael M. C. S., Valentova, Jaroslava V., Varella, Marco A. C., Corral-Frías, Nadia S., Frías-Armenta, Martha, Hatami, Javad, Monajem, Arash, Sharifian, MohammadHasan, Frohlich, Brooke, Lin, Hause, Inzlicht, Michael, Alaei, Ravin, Rule, Nicholas O., Lamm, Claus, Pronizius, Ekaterina, Voracek, Martin, Olsen, Jerome, Giolla, Erik Mac, Akgoz, Aysegul, Özdoğru, Asil A., Crawford, Matthew T., Bennett-Day, Brooke, Koehn, Monica A., Okan, Ceylan, Gill, Tripat, Miller, Jeremy K., Dunham, Yarrow, Yang, Xin, Alper, Sinan, Borras-Guevara, Martha Lucia, Cai, Sun Jun, Tiantian, Dong, Danvers, Alexander F., Feinberg, David R., Armstrong, Marie M., Gilboa-Schechtman, Eva, McCarthy, Randy J., Muñoz-Reyes, Jose Antonio, Polo, Pablo, Shiramazu, Victor K. M., Yan, Wen-Jing, Carvalho, Lilian, Forscher, Patrick S., Chartier, Christopher R., and Coles, Nicholas A.

Published
2021
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results