Deciding among models : a decision-theoretic view of model complexity
MetadataShow full item record
This research examines the trade-off between the cost of adding complexity to a model and the value added to the results within the context of decision-making. It seeks to determine how complex a model should be in order to fit it to the purpose at hand. The report begins with a discussion on general modeling theory and model complexity. It next considers the specific case of petroleum reservoir models and the existing research that has compared modeling results with model complexity levels. Finally, it presents original results applying Monte Carlo sampling to a drilling decision scenario and to a one-dimensional reservoir model where a cylindrical oil field is represented by different numbers of cells and the results compared.
Showing items related by title, author, creator and subject.
A joint multiple discrete continuous extreme value (MDCEV) model and multinomial logit model (MNL) for examining vehicle type/vintage, make/model and usage decisions of the household Sen, Sudeshna (2006)In this dissertation, we seek to contribute to the area of automobile demand modeling by developing a comprehensive econometric model to examine several dimensions of household vehicle holdings and usage decisions. In ...
Modeling residential self-selection in activity-travel behavior models : integrated models of multidimensional choice processes Pinjari, Abdul Rawoof (2008-08)The focus of transportation planning, until the past three decades or so, was to provide adequate transportation infrastructure supply to meet the mobility needs of the population. Over the past three decades, however, in ...
Multivariate multiple-membership random effects models : a demonstration and assessment of model estimation and fit Park, Sunyoung, Ph. D. in educational psychology (2018-04-27)The current dissertation, composed of two studies, focused on the models that handle several data structure complexities simultaneously. The first study introduced and evaluated Markov Chain Monte Carlo (MCMC) estimation ...