HT24
Cross-validation is a technique widely used in machine learning to evaluate the performance of a model. It provides insight into the model's performance on unseen data, helps avoid overfitting, and leads to the selection of models with better predictive ability. Despite its advantages, cross-validation is a computationally expensive method. However, in the field of pharmacometrics, the main metric for model evaluation is often the objective function value (likelihood-based approaches), as the goal is to find the best-fitting model for the data. Although a cross-validation method is not often used in pharmacometric model selection, it is sometimes used in covariate model building.
The project aims to evaluate cross-validation techniques in pharmacometric model selection, assessing its potential benefits and limitations compared to likelihood-based approaches.
The work will be carried out in Uppsala, however in a research group, that offers an international environment. The working language, including supervision, will be English.
Does this project sounds like something for you? Apply here or contact the supervisor (see below) or Maria Kjellsson (maria.kjellsson@farmaci.uu.se) if you have additional questions. We look forward to your application!
Farmaceutisk vetenskap
Farmakometri
Beräkningsstudie
Uppsala University
Uppsala
Thanakorn Vongjarudech, Elin Svensson
thanakorn.vongjarudech@farmaci.uu.se; elin.svensson@farmaci.uu.se
Institutionen för farmaci
Masterprogram i läkemedelsmodellering
Degree Project in Pharmaceutical Modeling within Pharmacometrics 45 c - 3FB029
45hp
1