Wyniki wyszukiwania - BazTech

Ograniczanie wyników

3 Zeszyty Naukowe. Włókiennictwo / Politechnika Łódzka

3 2000

Znaleziono wyników: 3

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: date analysis

Sortuj według:

Ogranicz wyniki do:

Exploration of univariate data in the textile metrology

Militky J., Meloun M.

Zeszyty Naukowe. Włókiennictwo / Politechnika Łódzka

2000

z. 58

165-173

The main aim of data analysis in textile metrology is the extraction of relevant information from measurements. This contribution is focused to most frequent task, i.e univariate data treatment based on the analysis of experimentally determined values Xi, i=1,...N. The system of exploratory data analysis based on concept of quantile estimation is proposed. The main part of this contribution is devoted to the description of several graphical tools for data summarization and exploration. Selected techniques are used for analysis of the strength of basalt fibers at small gauge length. The program package ADSTAT for realization of the above-mentioned techniques is briefly described.

Data based probability models in the textile metrology

Militky J., Meloun M.

Zeszyty Naukowe. Włókiennictwo / Politechnika Łódzka

2000

z. 58

155-164

The main aim of data analysis in textile metrology is the extraction of relevant information from measurements and creation of probability models. This contribution is focused to most frequent task, i.e univariate data treatment based on the analysis of experimentally determined values Xi, i=1,...N. The main part of this contribution is devoted to the description and camparison of data distribution with theoretical ones. Selected techniques are used for analysis of the strength of basalt fibers at small gauge length.

Data analysis for quality control in the textile branch

Militky J., Meloun M., Kupka K.

Zeszyty Naukowe. Włókiennictwo / Politechnika Łódzka

2000

z. 58

43-50

The Box Cox transformation are given for construction of asymmetric quantiles and control limits for control charts. It is shown on simulated and real data that the transformation may significantly improve applicability of control charts.