Contact: , read more about my interest at my NMBU homepage here



Below are a handful of suggested master project. If you have other suggestions, ideas or external partners so feel free to come with suggestions.


Erfaringsutveksling mellom norske havredyrkere

Hilde Vinje i samarbeid med Karstein Brøndbo og Norsk Havreforening

Havre er den nest viktigste kornarten i norsk kosthold etter hvete, og den tredje mest dyrka kornarten her i landet etter hvete og bygg. Hver nordmann spiser ca 7 kg havre i året, mens mesteparten går til dyrefôr. Hvert år dyrker ca. \(4000\) bønder havre, ofte i vekstskifte med andre kornslag, eng, potet eller grønnsaker.

I tillegg til valg av vekstskifte står bonden overfor flere valg i havredyrkinga, som valg av havresort, valg av gjødselstrategier, valg av plantevern og timingen av ulike tiltak. Det tar tid å bygge seg opp erfaringer, og det kan være nyttig å bruke andres erfaringer i tillegg til egne. Norsk Havreforening har de to siste årene samlet inn data fra over 300 norske havredyrkere for å se hvordan geografi, bondens egne valg og ulike problemer har påvirket avlingen. Datamaterialet er presentert i et regneark. Her kan bøndene gjøre utvalg i dataene og se hvordan enkelte valg har påvirket avlingene innenfor dette utvalget. Dette er gjort ved pivotering, plot og regresjoner i Excel.

Målet med studien er å bygge statistiske metoder inn i behandling og presentasjon av data, slik at innsamlede erfaringer kan gi brukerne bedre beslutningsgrunnlag. Det kan f.eks gjøres ved å tydeliggjøre samvariasjon mellom variabler eller ved å vise signifikans i påviste forskjeller.

Les mer her.

Master projects in statistical method based on teaching statistics

Hilde Vinje and Kathrine Frey Frøslie

It is a known fact that having a basic statistical understanding is important both for further education and to understand and cope with the world we live in. In an ongoing project we study how various cognitive factors, like brain physiology and personality theory, influence students’ learning capabilities in statistics. Several master projects based on data generated from this project will be relevant. The project will have main focus on the statistical methods and/or statistical applications to analyze data.




Meat yield in Norwegian slaughter pigs - covariance between ham, bacon, chops and other products

Lars Erik Gangsei og Hilde Vinje

Pork is the most widely eaten meat in Norway and the world accounting for approximately 39% and 36% of the total meat intake respectively. The average Norwegian consumed approximately 26 kg of pork meat in 2018. A carcass of pork is split in four main cuts; ham, loin, belly and shoulder. These main cuts are is processed further into a number of commercial products, bacon, chops, cured ham, trimmings, steaks, spareribs, ribs etc. The demand for these products vary between them, between marked and over time. Based on commercial and sustainability considerations, it is desirable that the composition of pig carcasses corresponds to consumer demands. We assume that pig carcass composition varies with overall fatness, carcass weight, gender and breed.

In order to do the analysis data from Animalia’s pilot plant might be utilized. Weight and content of fat, meat and bone for commercial parts are registered (see Gangsei et.al, 2018 for details) for approximately 1500 Norwegian slaughter pigs. Data might be analyzed by multivariate response linear regression models. Applying logit-inspired responses, and possibly utilizing the Matrix Normal distribution are possibilities that might improve interpretation capabilities for the model.

The aim of study is to evaluate how carcass composition vary with carcass weight, overall fat content, breed and gender. Furthermore, when these effects are accounted for we want to analyze the covariance between the relative sizes of different parts of pig carcasses. Answers too these questions are important for slaughterhouses, pig farmers, farmer organizations and breeding companies, in order to improve commercial yield and streamline resource utilization.


References:

Gangsei, L. E., Bjerke, F., Røe, M., & Alvseike, O. (2018). Monitoring lean meat percentage predictions from optical grading probes by a commercial cutting pattern. Meat science, 137, 98-105.