Wyniki wyszukiwania - Biblioteka Nauki

1

APPLICATION OF SELECTED STATISTICAL METHODS TO ICT USED IN PUBLIC PROCUREMENT SYSTEM IN POLAND

100%

Starzyńska W. , Wiktorowicz J.

Metody Ilościowe w Badaniach Ekonomicznych

|

2009

|

tom 10

|

nr 1

221-231

EN

The paper discusses the application of multivariate analysis tools to the development of e-procurement and to the determinants of the process using data derived from a survey of the awarding entities. In particular, statistical methods were used to measure, classify and model the development of computerized public procurement.

2

Macroeconomic factors in modelling the SMEs bankruptcy risk. The case of the Polish market

80%

Ptak-Chmielewska A. , Matuszyk A.

Econometrics. Ekonometria. Advances in Applied Data Analytics

|

2019

|

nr vol. 23 no. 3

40-49

EN

The last financial crisis affected the SMEs sector in different countries at different levels and strength. SMEs represent the backbone of the economy of every country. Therefore, they need bankruptcy prediction models easily adaptable to their characteristics. In our analysis we verified hypothesis: including information about macroeconomic conditions significantly increases the effectiveness of the bankruptcy model. The data set used in our research contained information about 1,138 SMEs. All information was taken from the financial statements covering the period 2002-2010. The sample included enterprises from sectors: industry, trade and services. Selected financial ratios were used to build the model and the macroeconomic variables were added: GDP, inflation, and the unemployment rate. Logistic regression as the research method was applied. In our study we showed that the incorporation of the macro variables improved the prediction of the SMEs bankruptcy risk.

3

Some Practical Problems of Multivariate Survival Analysis of Epidemiological Studies

80%

Jasiński B. , Kawalec E. , Kupść W. , Bogdan Jasiński E. K. J. U. , Witold Kupść - National Institute of Cardiology

Acta Universitatis Lodziensis. Folia Oeconomica

|

2002

|

tom 162

EN

In the epidemiological analysis of chronic diseases (most often cardiovascular or cancer) the main problem of interest is the estimation of the risk of death (or getting ill) related to set of characteristics called risk factors. For epidemiological studies typical features are: - large sample size (at least 1000 persons), - long follow up period for survival analysis (5 or more years), - large percentage of censored observations (patients who survive the whole time of study, more than 90%), - large number of registered risk factors. Some practical problems that concern the statistical analysis of the epidemiological data are following: - selection of the survival function model, - selection of the variables included into the model, - inclusion of interaction and/or higher order effect into the model. Some solutions of presented problems were applied to the Polish Part of Cardiovascular Diseases Prevention Program (Euro 8202). The program was conducted in 1976-1982 years with long follow up period concerning mortality till 1994 year. The program covered 8603 working men aged 40-59 years in two regions - Warsaw and South-Eastern Poland. Most of statistical analyses were performed on the basis of standard Statistical Analysis System (SAS) package.

PL

Jednym z głównych celów epidemiologicznych badań nad chorobami przewlekłymi (najczęściej układu krążenia lub nowotworowymi) jest oszacowanie ryzyka zachorowania lub zgonu w zależności od zespołu cech - czynników ryzyka. Badania epidemiologiczne charakteryzują się najczęściej następującymi własnościami: - duża liczebność próby, powyżej 1000 badanych; długi okres obserwacji badanych osób, ponad kilka lat; - wysoka frakcja (ok. 90%) osób, które przeżyły cały okres badania bez incydentu chorobowego, tzw. cenzorowanie administracyjne; - duża liczba czynników ryzyka rejestrowanych w badaniu. Analiza statystyczna badania epidemiologicznego wymaga, między innymi, rozwiązania następujących problemów: - wybór modelu funkcji oceniającej ryzyko, - selekcja badanych w modelu czynników ryzyka, - ocena wzajemnego oddziaływania (interakcji) badanych czynników i ocena nieliniowych efektów ich oddziaływania. Rozwiązanie przedstawionych zadań przeprowadzono na przykładzie analizy wyników Polskiego Programu Prewencji Chorób Układu Krążenia przeprowadzonego w latach 1976-1982, obejmującego 8603 mężczyzn zatrudnionych w zakładach pracy w dwóch regionach Polski -Warszawy i Polski Południowo-Wschodniej i rozszerzonego o obserwację postępującą w zakresie zgonu do roku 1994.

4

The Impact Of Previous Job Experience on Employment Odds in Szczecin

80%

Bieszk-Stolorz B. , Markowicz I.

Equilibrium. Quarterly Journal of Economics and Economic Policy

|

2012

|

tom 7

|

nr 4

EN

The aim of this article is to examine the impact of job experience on the odds. The studies which have been conducted by the authors so far focus on such determinants of finding a job by the unemployed as: gender, age and education. It has been confirmed that they are the features determining both the employment odds and the time devoted to seeking a job. The authors have presented a thesis that an unemployed person’s professional experience conditions affect the likelihood of their finding employment. Moreover, the odds are not the same in individual subgroups of a given community. The research tool used in the presented analysis is a model of logistic regression which, following the logit transformation, enables the researchers to determine the odds ratio. The odds ratio makes it possible to compare the employment odds of a person who declares previous employment experience with that of a person who has not been employed before. The authors examined the influence of previous job experience on employment odds in a given community as a whole and in individual subgroups divided by gender, age and education. Statistical data were obtained thanks to a long-term cooperation with the Poviat Labour Office in Szczecin. The analysed data covered 19 398 people who unregistered from the Poviat Labour Office in 2009.

5

Application of Logistic Regression for Firms Survival Analysis

80%

Markowicz I. , Solorz B. , Department of Econometrics and Statistics U. o. S.

Acta Universitatis Lodziensis. Folia Oeconomica

|

2009

|

tom 228

PL

Lata 90-te XX w. w Polsce to okres rozwoju przedsiębiorczości. Swoboda zakładania i prowadzenia działalności gospodarczej przyczyniła się do powstawania nowych, głównie małych, firm. Nie wszystkim przedsiębiorcom udaje się jednak utrzymać działalność w dłuższym czasie. Celem artykułu jest analiza szansy przetrwania firmy przez określony czas oraz różnic dla firm założonych w różnych latach. Zastosowano model regresji logistycznej dla dychotomicznej zmiennej zależnej. W artykule przedstawiono wyniki etapu badań prowadzonych w ramach projektu badawczego MNiSW N 111 011 31/1109.

EN

The 90's of the 20th century in Poland it's a period of economic development. Freedom o f establishing and carrying on a business caused rising many new, mostly small firms. Yet not all entrepreneurs were able to run their businesses for a longer period of time. The purpose of the paper is to analysis firms survival chance in a determined period o f time and to analysis differences between firms established in different years. Logistic regression model for dichotomous dependent variable will be used.

6

On Distance-Based Algorithms in Medical Application

80%

Misztal M. , Banach M. , Małgorzata Misztal C. o. S. M. , Maciej Banach C. o. C. a. C. S.

Acta Universitatis Lodziensis. Folia Oeconomica

|

2009

|

tom 228

PL

W badaniach medycznych do przewidywania przynależności pacjentów do jednej z wyróżnionych dwóch klas zwykle wykorzystuje się model regresji logistycznej. Algorytmy minimalnoodległościowe, takie jak np. algorytm najbliższego sąsiada, mimo ich prostoty i intuicyjnej interpretacji, są wykorzystywane bardzo rzadko. W referacie podjęto próbę zastosowania algorytmów opartych na odległościach (NN, k-NN, DB oraz k-NN Tree) do prognozowania wystąpienia migotania przedsionków wśród 300 pacjentów po zabiegu wymiany zastawki aortalnej.

EN

Logistic regression is the most popular method used to classify patients into 2 selected subgroups in medical research. Distance-based algorithms, such as nearest neighbor algorithm, simple and intuitive, are rarely used in practice. In the study some selected distance-based algorithms (NN, k-NN, DB and k-NN Tree) were applied to predict atrial fibrillation (AF) incidents among 300 patients with aortic valve defects, who underwent aortic valve replacement.

7

The employment of Łódź University tourism and recreation graduates: A sociological analysis

80%

Żulicki R.

Turyzm

|

2017

|

tom 27

|

nr 1

65-74

EN

This article is focused on selected aspects of the economic ‘fate’ of Tourism and Recreation graduates of the University of Łódź (UŁ). Its aim is to seek answers to the question: ‘What determines graduate employment?’ Surveys conducted by the Career Office of University of Łódź among graduates one year after graduation in 2014, 2015 and 2016 are the empirical basis. Tourism and Recreation graduates were compared with others from the Faculty of Geographical Sciences UŁ. The logistic regression technique was used to predict the status of graduate employment based on independent variables. The strongest predictors of graduate employment were structural and institutional characteristics. The quantitative results were interpreted in the context of the modern role of universities.

8

Production Potential and Income of Agricultural Holdings in Poland

80%

Strzelecka A. , Zawadzka D. , Kurdyś-Kujawska A.

Barometr Regionalny. Analizy i prognozy

|

2018

|

nr 3(53)

137-144

EN

The aim of this article is to assess the impact of production potential on the income obtained by agricultural holdings in Poland. The Farm Accountancy Data Network FADN data from 2015 was used to achieve this aim. Empirical verification of factors determining the production potential and their impact on agricultural holding income was carried out using the logistic regression model. The dependent variable was taken to be the probability that the agricultural holding would achieve annual family farm income exceeding the median value of PLN 46149,18. It was established that four variables had a statistically significant positive effect on the studied phenomenon: area of agricultural land, share of leased land in the agricultural land area, total labor expenditure and technical labor infrastructure. This means that an increase in the level of these factors increases the probability of obtaining an above median value of income by agricultural holdings.

PL

Celem artykułu jest ocena wpływu potencjału produkcyjnego na dochody uzyskiwane przez gospodarstwa rolne w Polsce. Do realizacji założonego celu wykorzystano dane Farm Accountancy Data Network (FADN) z 2015 roku. Weryfikację empiryczną czynników określających potencjał produkcyjny i ich wpływ na dochody gospodarstw rolnych przeprowadzono przy wykorzystaniu modelu regresji logistycznej. Za zmienną objaśnianą przyjęto prawdopodobieństwo osiągnięcia przez gospodarstwo rolne rocznego dochodu z rodzinnego gospodarstwa rolnego przekraczającego wartość 46 149,18 zł (mediana). Ustalono, że statystycznie istotny dodatni wpływ na badane zjawisko miały cztery zmienne: powierzchnia użytków rolnych, udział gruntów dzierżawionych w powierzchni UR, nakłady pracy ogółem oraz techniczne uzbrojenie pracy. Oznacza to, że wzrost poziomu tych czynników zwiększa prawdopodobieństwo uzyskania przez gospodarstwa rolne rocznego dochodu powyżej 46 149,18 zł.

9

Segregation of waste at source reduces the environmental hazards of municipal solid waste in Patna, India

80%

Singh A. , Raj P.

Archives of Environmental Protection

|

2018

|

tom Vol. 44, no. 4

96--110

EN

Though Municipal Solid Waste (MSW) is a worldwide problem, the collected wastes are dumped in open dumping at landfilling sites while the uncollected wastes remain strewn on the roadside, many-a-time clogging drainage. Such unsafe and inadequate management of MSW causes spread of bacteria, viruses, particulate matter, dioxins and other harmful pollutants in the surroundings and atmosphere. Hence, due to the repeated exposure of population to these pollutants can lead to serious health problems such as Diarrhea/Dysentery, Acute Respiratory Infection (ARI), and Asthma/Chronic Respiratory Diseases (CRD). Therefore, two-phase study included secondary data on diseases caused due to environmental pollution and primary data on MSW and lack of MSW management from 127 households in urban Patna, India. The random sampling method was used for collection of primary survey data, conducted during 2015–16 in selected areas of Patna. Logistic regression model odds ratios and their 95% confidence intervals were used to show the strength of the associations among segregation of wastes at source, segregation behavior, collection bins in the area, distance of collection bins from a residential area, and transportation of MSW. The ROC is a statistical technique to validate the logistic regression method that predicts the occurrence of an event through the comparison of probability picture of an event occurrence observed by probability and the predicted probability of the same event. The area under the ROC curve is up to 0.889 extent, which reveals that the ‘segregation of waste at source’ has a very strong scope to accomplish sustainable recycling at urban Patna in order to manage waste with the overall accuracy of 92.126%, which proves a better fi t logistic regression model. Hence, this paper concludes that ‘segregation of waste at source’ helps to attain sustainable recycling which would be the most viable approach to manage MSW in Patna and would eventually reduce environmental pollutants for the public health safety.

10

On Application of Logistic Regression to Mean Value Estimation in Two-Phase Sampling for Nonresponse

80%

Gamrot W. , University of Economics in Katowice D. o. S.

Acta Universitatis Lodziensis. Folia Oeconomica

|

2005

|

tom 194

PL

Wystąpienie niekompletności obserwacji badanej cechy w badaniu statystycznym zazwyczaj prowadzi do obciążenia uzyskanej oceny badanego parametru populacji. Jedna z technik stosowanych dla przeciwdziałania temu zjawisku opiera się na wykorzystaniu schematu losowania dwufazowego. Jako estymator wartości przeciętnej w populacji wykorzystuje się zazwyczaj kombinację liniową ocen wartości przeciętnej uzyskanych w pierwszym i drugim etapie (fazie) badania. W niniejszym referacie podjęto próbę zbadania własności alternatywnych strategii estymacji wykorzystujących schemat losowania dwufazowego, uwzględniających w konstrukcji estymatora oceny prawdopodobieństw uzyskania odpowiodzi od poszczególnych jednostek populacji uzyskane przy wykorzystaniu modelu regresji logistycznej. Porównanie własności wymienionych strategii przeprowadzono drogą symulacji komputerowej, przy wykorzystaniu danych uzyskanych w wybranych gminach powiatu Dąbrowa Tarnowska podczas spisu rolnego w roku 1996.

EN

The phenomenon of nonresponse in a sample survey usually leads to bias in estimates of population parameters. One of the techniques applied as a countermeasure for nonresponse is based on two-phase (or double) sampling. Usually a linear combination of mean value estimates obtained in both phases of the survey is used as an estimate of population mean value of the characteristic under study. In this paper alternative estimators for two-phase sampling scheme using estimates of response probabilities obtained on the basis of logistic regression model are considered. The results of Monte Carlo simulation study comparing the properties of these estimators are presented. In the simulations, the data from the Polish 1996 Agricultural Census were used.

11

A comparison of bayes discrimination, location model and logistic regression on the basis of real

80%

Ćwiklińska-Jurkowska M. , Jurkowski P.

Biocybernetics and Biomedical Engineering

|

1998

|

tom Vol. 18, no. 3-4

5-17

EN

The paper compares alternative methodologic approaches for two-group discrimination with mixed explanatory variables. The important problem is to select the best discriminant method for the model. For this purpose on the basis of the data connected with recognizing childhood asthma we compare functioning of different methods. We have applied various discriminant methods based on mixed quantitative and qualitative observations. The discrimination is performed on the grounds of all mixed variables and also on the basis of the most discriminating variables. The classical methods and nonparametric ones, such as kernel (with different kernel functions) and the nearest neighbour (with different distances), are employed. We have also applied the logistic discrimination and the location linear model of the first order. For the explored examples of mixed model - nonparametric kernel methods wit h different covariance matrices gave the best results.

12

Application of the Beneish Model on the Warsaw Stock Exchange

80%

Sylwestrzak M.

Journal of Banking and Financial Economics

|

2022

|

nr 2(18)

5-16

EN

This paper investigates irregularities in financial statements by applying the Beneish and Roxas models to Polish firms listed on the Warsaw Stock Exchange from 2015 to 2020. The total sample included 110 observations. The sample comprised companies that had received an adverse or disclaimer opinion by the auditors, but had not been fined by the Polish Financial Supervision Authority (KNF Board). The control firms were selected based on the industry as selected by the standard industrial classification code and on the financial year, with minimizing the difference in the size of total assets. The results indicate that the Roxas model revealed greater accuracy than the Beneish model on the tested sample. The use of logistic regression allowed a modification of the Beneish model to align it with the conditions of the Polish market. The modified Beneish model showed greater accuracy for the tested sample and companies fined by the KNF Board.

13

Techniques of nominal data analyses

80%

Stanisz A. , Broniatowska E.

Bio-Algorithms and Med-Systems

|

2012

|

tom Vol. 8, no. 2

237--254

EN

Several advanced techniques for statistical analysis of nominal data were discussed to show how interesting associations between examined variables can be obtained using correspondence analysis, logistic regression and log-linear models. All these techniques are introduced on the example of medical data connected with the patients being on the curative diet. The studied data concerns four nominal variables: breaking the diet, overcoming diseases requiring the curative diet, sex and age of patient. Applied analyses were used for searching which of factors mentioned above influence on breaking the diet. The most popular technique in this case, chi-squared test of independence, indicates that only the age and illnesses overcoming before are related with breaking the diet whereas the sex is factor which does not have any relationship with the diet breaking. However, the deeper analysis revealed that we can not omit this variable in our research. Application of more compound statistical methods show the importance of age and sex in breaking the curative diet in detail. Presented methodology can be successfully applied not only in medicine but to data coming from different branches of science as well.

14

ZNACZENIE ZASIŁKÓW DLA BEZROBOTNYCH DLA PRZEPŁYWÓW Z BEZROBOCIA DO ZATRUDNIENIA

80%

Gałecka-Burdziak E.

Polityka Społeczna

|

2011

|

tom 38

|

nr 9(450)

17-21

EN

Unemployment benefits are created to financially assist the unemployed while searching the job. The aim of this paper is the analysis of the selected determinants (among others: age, education, reservation wage, search effort) of the mentioned flow between unemployment and employment taking into account the influence of the unemployment benefits using Labour Force Survey data during the period 2007–2009. Analyzed period should constitute a relevant reference period taking into account changes in the size of the unemployment benefits introduced at 1st January, 2010.

15

THE ANALYSIS OF CHANCES OF YOUNG AND MIDDLE-AGED PEOPLE FOR HAVING A JOB USING BAYESIAN LOGISTIC REGRESSION MODEL

80%

Grzenda W.

Metody Ilościowe w Badaniach Ekonomicznych

|

2017

|

tom 18

|

nr 1

27-37

EN

The aim of this article is to analyze the chances of having a job using Bayesian logistic regression model. In this study both young and middle-aged people have been considered. The individual characteristics of economically active people have a significant impact on their labour market status. In this research the commonly studied set of features has been extended by adding the following characteristics: marital status, financial situation of the household, health assessment and the fact of living with parents in the case of young people. In this study, Bayesian logistic regression model has been used. The Bayesian approach enabled us to incorporate information from previous studies.

16

Prediction of changes in the tax burden of land plots with the use of multivariate statistical analysis methods

80%

Dmytrów K. , Gnat S.

Econometrics. Ekonometria. Advances in Applied Data Analytics

|

2019

|

nr vol. 23 no. 2

33-48

EN

It is believed that the ad valorem tax will increase fiscal burdens. In order to verify this statement, with the use of the Szczecin Algorithm of Real Estates Mass Appraisal, the land plots were appraised and the ad valorem tax was calculated. Next, a training set was sampled, for which the composite variable was calculated by means of three approaches: the TOPSIS method, the Generalised Distance Measure as the composite measure of development (GDM2), and the quasi-TOPSIS. They were the explanatory variables in the logistic regression model. Next, for the test set, changes of tax burden were forecasted. The aim of the research was to check the effectiveness of the presented approach for the estimation of the consequences of introducing the ad valorem tax. The results showed that all three approaches yielded similar results, but GDM2 was the best one. The main finding is that these approaches can be used in the prediction of changes in the tax burden of land plots.

17

Regresja logistyczna dla danych symbolicznych interwałowych

80%

Pełka M.

Econometrics. Ekonometria. Advances in Applied Data Analytics

|

2015

|

nr 2 (48)

44-52

EN

When dealing with real data situation we often have a binary (biomial, dichoto-mous) dependent variable. As the linear probability model is not such a good solution in such a situation there is a need to use nonlinear models. A quite good solution for such a sit-uation is the logistic regression model. The paper presents an adaptation of linear regression model when dealing with symbolic interval-valued variables. Four approaches poposed by de Souza et. al [2011] how to apply such variables are presented. In the empirical part re-sults obtained with the application of artificial and real data sets are shown. The best results are obtained for midpoint and bounds (joint estimation) methods.

18

Decision-making process in colon disease and crohn’s disease treatment

80%

Dardzińska A. , Kasperczuk A.

Acta Mechanica et Automatica

|

2018

|

tom Vol. 12, no. 3

227--231

EN

The article presents the process of building a logistic regression model, which aims to support the decision-making process in medicine. Currently, there is no precise diagnosis for ulcerative colitis (UC) and Crohn's disease (CD). Specialist physicians must exclude many other diseases occurring in the colon. The first goal of this study is a retrospective analysis of medical data of patients hospitalized in the Department of Gastroenterology and Internal Diseases and finding the symptoms differentiating the two analyzed diseases. The second goal is to build a system that clearly points to UC or CD, which shortens the time of diagnosis and facilitates the treatment of patients. The work focuses on building a model that can be the basis for the construction of classifiers, which are one of the basic elements in the medical recommendation system. The developed logistic regression model will define the probability of the disease and will indicate the statistically significant changes that affect the onset of the disease. The value of probability will be one of the main reasons for the decision.

19

A comparison of regularization techniques in the classification of handwritten digits

80%

Klimaszewski J.

Journal of Theoretical and Applied Computer Science

|

2015

|

tom Vol. 9, nr 4

3--7

EN

If dataset is relatively small (e.g. number of samples is less than number of features) or samples are distorted by noise, regularized models built on that dataset often give better results than unregularized models. When problem is ill-conditioned, regularizaton is necessary in order to find solution. For data where neighbouring values are correlated (like in images or time series), not only individual weights, but also differences between them may be penalized in the model. This paper presents results of the experiment, in which several types of regularization (l2, l1, penalized differences) and their combinations were used in fitting logistic regression model (trained using one-vs.-rest strategy) to find which one of them works the best for various sizes of training set. Data used in the experiment came from MNIST dataset, which is publicly available.

20

Statistical models in enterprises default risk assessment – an example of application

80%

Ptak-Chmielewska A. , Kuleta P.

Econometrics. Ekonometria. Advances in Applied Data Analytics

|

2018

|

nr vol. 22 no. 1

94-106

EN

Default risk assessment is crucial in the banking activity. Different models were developed in the literature using the discriminant analysis, logistic regression and data mining techniques. In this paper the logistic regression was applied to verify models proposed by R. Jagiełło for different sectors. As an alternative, the logistic regression model with the nominal variable SECTOR was applied on the pooled sample of enterprises. The dynamic approach using the Cox regression survival model was estimated. Including the nominal variable SECTOR only slightly increases the predictive power of the model (in the case of “defaults”). The predictive power of the Cox regression model is lower, the only advantage is the higher accuracy classification in the case of “defaulted” enterprises.