Wyniki wyszukiwania - BazTech

1

The critical path method in estimating project duration

Nafkha R., Wiliński A.

Information Systems in Management

|

2016

|

Vol. 5, No. 1

78--87

EN

Undoubtedly, the most important criterion for assessing any project success is to achieve the planned main objective within scheduled time, under the assumed costs and corresponding to established project quality parameters. The article presents critical path method, that aims to determine project schedule which ensure an implementation shortest time. Ultimately, however the optimal project duration is designated after solving human resource deficiencies or conflicts occurring in the generated schedule.

2

A study on the effectiveness of investment strategy based on the concept of pivot points levels using Matthews criterion

Wiliński A., Nyczaj T., Bera A., Błaszyński P.

Journal of Theoretical and Applied Computer Science

|

2013

|

Vol. 7, nr 4

42--55

EN

In this work, the possibility of assessing traditional investment strategy based on the pivot points for using with other than the commonly used criterion is examined. The authors attempted to apply the Matthews Correlation Coefficient (further reffered as MCC) criterion based on a confusion matrix when assessing the strategy to include more factors than the traditional criteria (such as profit, profit vs. Risk, Sharpe ratio, Calmar ratio) and to express these factors by one number. The criterion based on a confusion matrix is, in authors beliefs, unique in this application and gives a fairly valuable estimation of trading strategy. An example of several strategies tested on EURUSD 1h time series in selected intervals in the years 2012-2013 is considered. Among these strategies there is a simple strategy based on the concept of pivot points levels and more complex derivative strategies, based on the vector of optimized values of certain parameters. These strategies are evaluated using both traditional criteria and modification of MCC proposed by the authors.

3

Computer models of similarity in musical pieces based on the multi-dimensional patterns

Mazurowski Ł., Wiliński A.

Journal of Applied Computer Science

|

2012

|

Vol. 20, nr 2

79-94

EN

This paper describes the problem of recognizing similarities in musical pieces in order to cluster and classify them with particular reference to the files stored according to the MIDI standard. The analysis of the similarity between artificially generated musical pieces to those that have been composed by a man which is carried out in order not to infringe copyrights to the existing pieces is the area of further use of the method presented. The article presents different existing methodological approaches and proposes the use of histograms of selected parameters of musical sound as a mechanism of aggregation of musical clusters potentially belonging to one group of similar musical pieces.

4

Ensemble of data mining methods for gene ranking

Wiliński A., Osowski S.

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2012

|

Vol. 60, nr 3

461-470

EN

The paper presents the ensemble of data mining methods for discovering the most important genes and gene sequences generated by the gene expression arrays, responsible for the recognition of a particular type of cancer. The analyzed methods include the correlation of the feature with a class, application of the statistical hypotheses, the Fisher measure of discrimination and application of the linear Support Vector Machine for characterization of the discrimination ability of the features. In the first step of ranking we apply each method individually, choosing the genes most often selected in the cross validation of the available data set. In the next step we combine the results of different selection methods together and once again choose the genes most frequently appearing in the selected sets. On the basis of this we form the final ranking of the genes. The most important genes form the input information delivered to the Support Vector Machine (SVM) classifier, responsible for the final recognition of tumor from non-tumor data. Different forms of checking the correctness of the proposed ranking procedure have been applied. The first one is relied on mapping the distribution of selected genes on the two-coordinate system formed by two most important principal components of the PCA transformation and applying the cluster quality measures. The other one depicts the results in the graphical form by presenting the gene expressions in the form of pixel intensity for the available data. The final confirmation of the quality of the proposed ranking method are the classification results of recognition of the cancer cases from the non-cancer (normal) ones, performed using the Gaussian kernel SVM. The results of selection of the most significant genes used by the SVM for recognition of the prostate cancer cases from normal cases have confirmed a good accuracy of results. The presented methodology is of potential use for practical application in bioinformatics.

5

Poszukiwanie twarzy i rozpoznawanie emocji w scenach zmiennych

Wosiak S., Buda K., Wiliński A.

Metody Informatyki Stosowanej

|

2011

|

nr 5

147-157

EN

This article focuses on the problems encountered when using face observation and emotion recognition for the purposes of identification and also classification of specific emotions. The identification of emotions are particularly difficult, especially within varying scenes. In this paper we review the main methods of the identification of certain emotions. The authors present the results of their research analysis in tracking changes of the emotions in the selected software packages.

6

The management of prediction method in the system of investment decisions making

Wiliński A., Rejer I.

Management and Production Engineering Review

|

2011

|

Vol. 2, No. 2

46-52

EN

The aim of the paper is to find a method of using prediction rules in time series in such a way to maximize the profit considering the risk. To deal with this task, a regression approach to prediction was chosen. Hence, the paper refers to relation between autoregression of a chosen time series and investment strategies. The time series under consideration is the most important polish financial instrument, a future contract on WIG20. Usually, it is rather easy to prove statistically that the autoregression of a single time series cannot be considered as an effective method for forecasting WIG20 quotations for investment purpose. However, the authors find the relation between the autoregression (and also multi-regression) and real future values of WIG20 which can be the source of effective strategies. The paper presents both - the theoretical description of the proposed strategies and results of their application for monthly data of WIG20, unemployment rate and money supply in Poland (data from years 1995-2007).

7

Metoda hierarchiczna optymalnej selekcji genów do rozpoznania nowotworów na podstawie macierzy ekspresji genów

Wiliński A., Osowski S.

Przegląd Elektrotechniczny

|

2009

|

R. 85, nr 4

50-52

PL

Praca przedstawia hierarchiczne podejście do selekcji genów odpowiedzialnych za choroby nowotworowe. Metoda składa się z dwu etapów. W pierwszym etapie zastosowano 8 różnych metod wartościowania genów według ich zdolności rozpoznawczej, w tym 2 metod opartych na liniowej sieci SVM, dyskryminancie Fishera, analizie korelacyjnej danych oraz zastosowaniu hipotez statystycznych, (3 odmiany metody Kołmogorowa-Smirnowa oraz test Wilcoxona). Na podstawie statystycznych wyników selekcji 100 najlepszych genów wyselekcjonowanych przy użyciu każdej metody w drugim etapie przetwarzania poszukuje się cech wspólnych, które traktuje się jako cechy optymalne, najlepiej różnicujące próbki danych należących do różnych klas nowotworowych. W pracy skoncentrowano się na wynikach eksperymentów numerycznych i ich analizie dla trzech przypadków nowotworów: białaczka, nowotwór prostaty i płuc. Pokazano, że zaproponowane podejście pozwala uzyskać dobre wyniki separacji różnych rodzajów nowotworów, widoczne zarówno na obrazie graficznym rozkładu macierzy ekspresji jak i w miarach numerycznych jakości separacji.

EN

The paper proposes the hierarchical approach to the selection of the optimal set of genes for cancer recognition on the basis of the gene expression microarray. In the first stage 8 different methods of gene selection are applied to the microarray of gene expression. They include the application of linear Support Vector Machine, the Fisher discriminant ratio, the correlation analysis and statistical hypothesis tests (Kolmogorov-Smirnov, Wilcoxon-Mann-Whitney). On the basis of statistical results of each selection method 100 most discriminative genes (the genes most often appearing in the selected set) are selected first. Then in the second stage the genes selected by all methods are compared. Only the genes discriminated simultaneously by all selected methods are chosen. In this way small number of the genes associated with the appropriate cancer type is selected. The numerical experiments performed for different types of cancer (prostate, lung cancer, leukemia) have proved the efficiency of the proposed approach. The PCA distribution of data and the distance measures associated with PCA have shown that the selected genes discriminate different cancer types very well. Also the graphical representation of the considered data show significant improvement of the recognition ability of the selected genes.

8

Strategia inwestowania na kontraktach terminowych na WIG20 oparta na koncepcji Saidenberga

Wiliński A.

Metody Informatyki Stosowanej

|

2009

|

nr 2 (19)

169-176

PL

Celem pracy jest przetestowanie na danych historycznych pewnej, istotnie zmodyfikowanej przez autora, koncepcji zaprezentowanej przez amerykańskiego eksperta rynków finansowych Richarda Saidenberga podczas wywiadu z innym znanym praktykiem giełdowym Joe Krutsingerem, opublikowanego w pracy tego ostatniego [10]. Koncepcja ta należy do grupy tzw. prostych reguł analizy technicznej, wyekstrahowanych spośród powtarzalnych wzorców w szeregach czasowych instrumentów finansowych. Najczęściej do reguł tych zalicza się grupy lub sekwencje warunków wykorzystujących rozmaite wskaźniki oparte na średnich kroczących, różnicach średnich, ich pierwszych pochodnych, odchyleniach standardowych (np. związanych z istotną w niniejszej pracy wstęgą Bollingera) czy punktach zwrotnych (pivot points). Friesen, Weller i Dunham [8] twierdzą, że metody te były przez wiele lat lekceważone przez środowisko akademickie pomimo częstego stosowania ich przez praktyków giełdowych. Cai B., Cai C. i Keasey [3] zauważyli skuteczność zarówno predykcyjną jak i inwestycyjną najprostszych reguł opartych na różnicach średnich kroczących i ich pochodnych, a także poziomach zmiany kierunku ruchu ceny (punktach zwrotnych). Tian, Wan i Guo [17] stwierdzili efektywność niektórych prostych reguł na jednych rynkach i całkowitą ich nieprzydatność na innych. Rozróżniali pod tym względem rynki dojrzałe i rosnące (np. amerykański i w szczególności w opozycji do niego – rynek chiński). Do przekonanych zwolenników prostych reguł należą także tacy badacze jak Brock i Lakonishok [2], LeBaron [2,11], Gencay [9], czy też wielu wybitnych praktyków takich jak najważniejszy w niniejszej pracy – Richard Saidenberg, Joe Krutsinger, Larry Williams, Joseph DiNapoli czy Michael Connor [10].

EN

The paper presents a trading strategy, tested using historical data, aimed at trading WTG20- based futures, inspired by a concept proposed by Richard Saidenberg. The strategy is based on simple technical analysis rules, can be also treated as a pattern recognition method. A primary goal consists in finding a pattern allowing for detecting a reversal point, basing on the lagged Bollinger band. The strategy has been verified using several years long time series of WIG 20-based futures quotes. Satisfactory results were obtained, considering prediction accuracy and practical usability.

9

Koncepcja generowania utworu muzycznego przez maszynę – akt sztucznej twórczości opartej na algorytmach genetycznych

Mazurowski Ł., Wiliński A.

Metody Informatyki Stosowanej

|

2008

|

nr 3 (Tom 16)

203--212

EN

The paper presents the examples of melodic line generated in a completely automated way, without human participation, using genetic algorithms, authors’ concepts and implementations. Fitness functions used in the algorithm consider several perceptual aspects of esthetical evaluation of musical creations. Hence the following do not occur in the generated individuals-compositions: dissonance, excessive frequency range, excessive tonal volatility. The artistic value of compositions was not assessed, since this could be accomplished only through statistical listeners’ polling, nevertheless the compositions do not irritate, staying in compliance with human esthetical paradigm.

10

Badania możliwości wyboru optymalnej zmiennej obserwowanej jako zmiennej predykcyjnej w zbiorze danych o rynkach finansowych

Wiliński A., Czapiewski P.

Metody Informatyki Stosowanej

|

2008

|

nr 1 (Tom 13)

159-174

EN

The paper presents the results of computer simulations performed using the historical quotes on several securities (WIG20, S&P500, Dow Jones, DAX, EUR/USD, gold, oil, etc.) in order to analyse the possibility of finding such variables, that can be explained in terms of the others better, than the rest. It is assumed, that the ultimate goal of every investment strategy is finding the opportunity of gaining a financial profit (always considering the risk). Such opportunity is being sought by investigating the possibility of using each variable (each security) in turn as the one to be predicted. In order to reach that goal, authors use several variants of one of the algorithms belonging to the. Group Method of Data Handling (GMDH), namely the combinatorial algorithm. The results reveal some interesting features of regression models, indicating the prospect of further applications of the method.

11

Model częstotliwości podstawowej traktu głosowego zbudowany metodą grupowania argumentów (GMDH)

Wiliński A., Samborska-Owczarek A.

Metody Informatyki Stosowanej

|

2007

|

nr 2(Tom 12)

189-202

EN

This paper concerns the problem of parameters estimation for a certain model, aiming at the approximation of output variable at the acceptable accuracy level. What distinguishes the way this common scientific task is here dealt with, is the usage of GMDH - Group Method of Data Handling (or more specifically the GMDH-based algorithm developed by the authors), which allows for simultaneous determination of both the structure and numerical characteristics of the model. The feature space under consideration is the matrix of repetitively observed attributes, describing the physical characteristics of voice samples, collected in order to determine the frequency of laryngeal tone for the purpose of medical diagnosis.