Parameter estimation in a Myeloproliferative neoplasm model

Maria Nicoline Nørr, Jeannie Bøg Rasmussen & Mehmet Temizsoy

Studenteropgave: Kandidatprojekt

Abstrakt

The aim of thisprojectwork was toestimate the parameter values in a given modelwhichcontainsfourcouplednon-linearordinarydifferentialequations, and access if the given model of the leukemia Myeloproliferative neoplasm disease is a good representative of the disease. This was performed with the given model and a small set of data from Roskilde Hospital. The parameter values in the model were initially guessed from the data and modelled graphically. A sensitivity analysis was conducted to rank the most sensitive parameters which was optimized using the Nelder-Mead optimization. The optimized parameter values from the hospital data were then used to create a set of pseudo data which were used to see if the model could reproduce the known parameter values. The optimization was conducted with the same parameter values, as for the hospital data, besides from the one parameter that was optimized. The value of this parameter was randomly chosen. The model was able to reproduce the parameter value back to the optimizedparametervaluesfromthehospitaldatauntiltherandomlychosen parameter value is within certain boundaries.

UddannelserMatematik, (Bachelor/kandidatuddannelse) Kandidat
SprogEngelsk
Udgivelsesdato7 apr. 2017
Antal sider103
VejledereJohnny T. Ottesen