Wyniki wyszukiwania - Biblioteka Nauki

1

Krzywe operacyjno-charakterystyczne w ewaluacyjnych badaniach marketingowych

100%

Sagan A.

Zeszyty Naukowe Uniwersytetu Ekonomicznego w Krakowie

|

2011

|

nr 864

5-17

EN

The articles examines the characteristics and use of the receiver operator characteristic (ROC) in research on marketing phenomena. This approach is an extremely popular tool for assessing prediction accuracy in medical research and signal detection, and is being ever more widely used in social and marketing research. The article presents the main indicators of classification accuracy for contingency tables (2×2) and the principles governing the use of ROC curves. The use of these curves is illustrated with an analysis of the accuracy of the choice of car (new or used) based on the structure of customer benefits.

2

Comparison of methods used for filling partially unobserved contingency tables

100%

Kot M. , Kamiński B.

Przegląd Statystyczny

|

2021

|

tom 68

|

nr 4

1-20

EN

In this article, we investigate contingency tables where the entries containing small counts are unknown for data privacy reasons. We propose and test two competitive methods for estimating the unknown entries: our modification of the Iterative Proportional Fitting Procedure (IPFP), and one of the Monte Carlo Markov Chain methods called Shake-and-Bake. We use simulation experiments to test these methods in terms of time complexity and the accuracy of searching the space of feasible solutions. To simplify the estimation procedure, we propose to pre-process partially unknown contingency tables with simple heuristics and dimensionality-reduction techniques to find and fill all trivial entries. Our results demonstrate that if the number of missing cells is not very large, the pre-processing is often enough to find fillings for the unknown values in contingency tables. In the cases where simple heuristics are insufficient, the Shake-and-Bake technique outperforms the modified IPFP in terms of time complexity and the accuracy of searching the space of feasible solutions.

3

ON THE METHOD OF COMPARING STRUCTURES BASED ON THE DATA IN THE CONTINGENCY TABLES

100%

Polko D.

Acta Universitatis Lodziensis. Folia Oeconomica

|

2014

|

tom 3

|

nr 302

EN

Comparison of populations is one of the most important problems in statistics. The most common comparisons apply to two populations, but comparisons of k populations, where k > 2 are also carried out. Parametric methods allow to compare the means, variances or proportions. The non-parametric methods allow to compare the distributions of two or more populations. The problem of comparison structures based on data in contingency tables is analyzed in the paper. The permutation tests were applied in the multivariate nominal data structure comparison.

4

Nonparametric Versus Parametric Reasoning Based on 22 Contingency Tables

88%

Sulewski P.

Computational Methods in Science and Technology

|

2018

|

tom Vol. 24, No. 2

143--153

EN

This paper proposes scenarios of generating contingency tables (CTs) with the probability flow parameter (PFP). It also defines measures of untruthfulness of H0 that involve PFP for all proposed scenarios. This paper is an attempt to replace a nonparametric statistical inference method by the parametric one. The paper applies the maximum likelihood method to estimate PFP and presents instructions to generate CTs by means of the bar method. The Monte Carlo method is used to carry out computer simulations.

5

The Problem of Zero Cells in the Analysis of Contingency Tables

63%

Brzezińska J.

Zeszyty Naukowe Uniwersytetu Ekonomicznego w Krakowie

|

2015

|

nr 5(941)

49-61

EN

Log-linear analysis is a statistical tool used to analyse the independence of categorical data in contingency tables. With this method, any number of nominal or ordinal variables can be analysed: interactions can be included in the model, various types of association can be analysed, and the analysis provides a formal model equation. Although log-linear analysis is a versatile statistical method, there are some limitations in using it due to zero cells. Zero cells in contingency table are of two types: fixed (structural) and sampling zeros. Fixed zeros occur when it is impossible to observe values for certain combinations of the variable. Sampling zeros are due to sampling variations and the relatively small size of the sample when compared with a large number of cells. In the paper several options will be presented for how to deal with zero cells in a table. All calculations will be conducted in R

PL

Analiza logarytmiczno-liniowa jest metodą badania zależności pomiędzy zmiennymi niemetrycznymi w tablicy kontyngencji, która pozwala analizować dowolną liczbę zmiennych nominalnych i porządkowych. Pomimo że jest ona uniwersalną metodą analizy zmiennych niemetrycznych, występują jednak pewne ograniczenia w jej stosowaniu ze względu na zerowe liczebności. Zera występujące w tablicy mogą być dwojakiego rodzaju: strukturalne lub związane ze schematem losowania. Zera strukturalne pojawiają się wtedy, gdy nie jest możliwa obserwacja kategorii zmiennej, a zera związane ze schematem losowania występują w małych próbach i znikają, gdy próba zostanie odpowiednio zwiększona. W artykule zaprezentowano sposoby radzenia sobie z zerowymi liczebnościami w tablicy kontyngencji. Wszystkie obliczenia przeprowadzono w programie R

6

Membership function - ARTMAP neural networks

63%

Sinčák P. , Hric M. , Vaščák J.

TASK Quarterly : scientific bulletin of Academic Computer Centre in Gdansk

|

2003

|

tom Vol. 7, No 1

43-52

EN

The project deals with the application of computational intelligence (CI) tools for multispectral image classification. Pattern Recognition scheme is a global approach where the classification part is playing an important role to achieve the highest classification accuracy. Multispectral images are data mainly used in remote sensing and this kind of classification is very difficult to assess the accuracy of classification results. There is a feedback problem in adjusting the parts of pattern recognition scheme. Precise classification accuracy assessment is almost impossible to obtain, being an extremely laborious procedure. The paper presents simple neural networks for multispectral image classification, ARTMAP-like neural networks as more sophisticated tools for classification, and a modular approach to achieve the highest classification accuracy of multispectral images. There is a strong link to advances in computer technology, which gives much better conditions for modelling more sophisticated classifiers for multispectral images.