Statistička analiza stabilnosti

master's thesis

• Author: Mia Franić
• Mentor: Vesna Lužar-Stiffler (mentor)
• Committee member: Vesna Lužar-Stiffler (predsjednik povjerenstva)
• Committee member: Pavle Goldstein (član povjerenstva)
• Committee member: Vedran Krčadinac (član povjerenstva)
• Committee member: Goran Muić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Mathematics); Zagreb
• Defense date and country: 2017-09-27, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Mathematics

Abstract

U ovom diplomskom radu opisana je metoda statističke analize stabilnosti farmaceutskih proizvoda, te su proučene matematičke metode linearne regresije, analize varijance i analize kovarijance. Također proučena je i metoda dugoročne analize stabilnosti s fiksnim efektima, a ukratko je obrađena metoda dugoročne analize stabilnosti sa slučajnim efektima. Proveden je Monte Carlo eksperiment u kojem je ispitana robusnost statističke analize stabilnosti s fiksnim efektima s obzirom na odstupanje slučajne pogreške od tražene distribucije variranjem standardne devijacije, te koeficijenata asimetrije i spljoštenosti. Eksperiment je pokazao da metoda nije robusna s obzirom na testirane parametre i da je osjetljiva na uzorke malih veličina. Na kraju je na rezultatima eksperimenta napravljena analiza distribucija roka trajanja s obzirom na parametre varirane u eksperimentu.

Abstract (english)

In this thesis method of statistical stability analysis of pharmaceutical products was studied. Mathematical methods of linear regression, variance and covariance analysis were discussed. Long term stability testing was studied with emphasis on stability testing with fixed effects and a brief overview of stability testing with random effects. Subsequently a Monte Carlo experiment was conducted, its aim was to test how robust statistical stability analysis with fixed effect is in regard of distribution of random error by variating standard deviation, coefficients of skewness and kurtosis. Experiment showed that described method is not robust and that method is quite sensitive to small sample sizes. At the end, the distribution of shelf-life was examined.

• Keywords: statistička analiza stabilnosti farmaceutskih proizvoda; linearna regresija; analiza varijance; analiza kovarijance; Monte Carlo eksperiment
• Keywords (english): statistical stability analysis of pharmaceutical products; linear regression; variance analysis; covariance analysis; Monte Carlo experiment
• Language: croatian
• URN:NBN: urn:nbn:hr:217:406344
• Study programme: Title: Mathematical Statistics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra matematike (magistar/magistra matematike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2018-01-26 12:14:42