Wyniki wyszukiwania - BazTech

1

Strategie treningu neuronowego estymatora częstotliwości tonu krtaniowego z użyciem generatora syntetycznych samogłosek

Blok Marek, Banaś Jan, Pietrołaj Mariusz

Przegląd Telekomunikacyjny + Wiadomości Telekomunikacyjne

|

2022

|

nr 4

604--607

PL

W wielu zastosowaniach telekomunikacyjnych pojawia się problem przetwarzania lub analizy sygnału mowy, w ramach którego, często w obszarze podstawowych algorytmów, stosuje się estymator częstotliwości tonu krtaniowego. Estymator rozpatrywany w tej pracy bazuje na neuronowym klasyfikatorze podejmującym decyzje na podstawie częstotliwości oraz mocy chwilowej wyznaczanych w podpasmach analizowanego sygnału mowy. W pracy rozważamy problematykę treningu tego estymatora, gdy trening odbywa się z użyciem sygnałów generowanych syntetycznie.

EN

In many telecommunication applications there is a need for a speech signal processing or analysis, within which the pitch tone frequency estimator is one of the common basic algorithms. The estimator considered in this paper is based on a neural classifier, whose decisions are driven by the instantaneous frequency and power determined in the sub-bands of the analyzed speech signal. In the paper, we consider the problems of selecting a training strategy for this estimator, when training is carried out with synthetically generated vowels.

2

Adaptive Differential Evolution with Elite Opposition-Based Learning and its Application to Training Artificial Neural Networks

Choi Tae Jong, Lee Jong-Hyun, Youn Hee Yong, Ahn Chang Wook

Fundamenta Informaticae

|

2019

|

Vol. 164, nr 2-3

227--242

EN

Differential Evolution (DE) algorithm is one of the popular evolutionary algorithms that is designed to find a global optimum on multi-dimensional continuous problems. In this paper, we propose a new variant of DE algorithm by combining a self-adaptive DE algorithm called dynNP-DE with Elite Opposition-Based Learning (EOBL) scheme. Since dynNP-DE algorithm uses a small number of population size in the later of the search process, the population diversity becomes low, and therefore premature convergence may occur. We have therefore extended an OBL scheme to dynNP-DE algorithm to overcome this shortcoming and improve the optimization performance. By combining EOBL scheme to dynNP-DE algorithm, the population diversity can be supplemented because not only the information of individuals but also their opposition information can be utilized. We measured the optimization performance of the proposed algorithm on CEC 2005 benchmark problems and breast cancer detection, which is a research field that has recently attracted a lot of attention. It was verified that the proposed algorithm could find better solutions than five state-of-the-art DE algorithms.

3

Analiza możliwości zastosowania sieci neuronowych do modelowania wartości kapitału społecznego w firmach IT

Siderska J.

Ekonomia i Zarządzanie

|

2013

|

Vol. 5, no. 1

84--97

PL

Celem artykułu jest przeanalizowanie możliwości zastosowania sztucznej sieci neuronowej do modelowania wartości kapitału społecznego w przedsiębiorstwach z sektora IT. Dane do badań zostały zaczerpnięte z bilansów finansowych publikowanych przez firmę Prokom Software SA i Asseco Poland SA na koniec każdego kwartału w latach 2005-2009. Niezbędnych obliczeń wartości kapitału społecznego firm dokonano z wykorzystaniem równania fundamentalnego. Uczenie sztucznej sieci neuronowej przeprowadzono w generatorze Statistica Neural Network 10.0. Badania pozwoliły na opracowanie szkieletu modelu odzwierciedlającego zależności pomiędzy zmiennymi wejściowymi (objaśniającymi), a zmienną wyjściową (objaśnianą), czyli wartością kapitału społecznego.

EN

The purpose of this article is to analyze the possibility of using an artificial neural network to model the value of social capital in companies from IT market. Training data were collected and calculated on the basis of financial report published quarterly by Prokom Software SA and Asseco Poland SA. The values of social capital in those companies were calculated by Fundamental Equation. The artificial neural network had been trained in STATISTICA Automated Neural Network 10.0 (SANN). The research contributed to developing the draft of the model for estimating the relationships among seven input variables and output variable, i.e. the value of social capital.

4

The interference spectrum extraction of a gait characteristics data record

Chandzlik S., Piecha J.

Journal of Medical Informatics & Technologies

|

2004

|

Vol. 7

KB23--32

EN

The paper shows several aspects of the gait data record analysis describing neurological diseases. The diagnosis of the gait abnormalities concerns interferences level of the patient physiological records. The disease source and level can be classified by the relevant interference functions. These functions were used for artificial records creation to multiply the necessary set of data needed for neural network training.

5

Modelling the data record of a patient walk by langrange-polynomial method

Chandzlik S., Piecha J.

Journal of Medical Informatics & Technologies

|

2002

|

Vol. 3

MI143--152

EN

Various options available in PSW footprint and walking characteristics measuring equipment [6], [7], give the user many aims in putting diagnosis. A Conclusion-Making Unit (CMU) that has been described in this paper supports the diagnosis automation procedures. Due to simplifying the CMU training process some affords in a field of the input record length reduction have been undertaken. The paper describes an analytical method of the data record description that allows converting discrete data samples into continuous function. This way a redigitalisation of the record can be done, where sampling period is matched with the walk length. This normalization allows reducing the data record length used for fast training of the CMU.

6

A patient walk-data-record modelling using a spline interpolation method

Chandzlik S., Piecha J.

Journal of Medical Informatics & Technologies

|

2002

|

Vol. 3

MI153--159

EN

The paper describes an analytical method of data record description that allows converting samples of discrete data record into continuous function. This operation allows re-sampling the data record with a sampling rate that is adequate to step duration. The record length is limited to an efficient size for training the Conclusion-Making Unit (CMU). Various options available in the PSW equipment [6], [7] give the user many aims in putting diagnosis anyhow, due to simplification of the CMU training process several methods for data records modifications are considered.