A multicriteria approach to model specification and estimation

Abstract

In decision theory, incommensurabilities among con icting decision criteria are typically handled by m ulticriteria optimization methods such a s P areto e ciency and meanvariance analysis. In econometrics and statistics, where con icting model criteria replace con icting decision criteria, probability assessments are routinely used to transform disparate model discrepancy terms into apparently commensurable quantities. This tactic has both strengths and weaknesses. On the plus side, it permits the construction of a single real-valued measure of theory and data incompatibility in the form of a likelihood function or a posterior probability distribution. On the minus side, the amalgamation of conceptually distinct model discrepancy terms into a single real-valued incompatibility measure can make it di cult to untangle the true source of any diagnosed model speci cation problem. This paper discusses recent theoretical and empirical work on a multicriteria \ exible least squares" (FLS) approach to model speci cation and estimation. The basic FLS objective is to determine the \cost-e cient frontier," that is, the set of estimates that are minimally incompatible with a speci ed set of model criteria. The relation of this work to previous work in econometrics, statistics, and systems science is also clari ed.

Time‐Varying Flexible Least Squares for Thermal Desorption of Gases

“…Ordinary linear least squares provides a single set of parameters for each evolution rate curve, and so does not provide information on time‐varying effects such as lateral interactions between absorbed species , mass and heat transfer processes , and surface heterogeneity and restructuring . FLS–TVLR is used to exactly simulate each data set by varying each of the parameters at each temperature. The resultant distribution provides accurate information on the heterogeneity of the gas‐surface reactions.…”

An objective function for simulation based inference on exchange rate data

“…Thus, performing the validation analysis moment by moment along the lines described above, is not feasible. Instead, validation has to be based either on the joint distribution of all moments with regard to some multicriteria approach (Kalaba and Tesfatsion 1996) or on a scalar valued objective function depending on all moments. The latter will be the approach discussed in the next subsection in the context of estimation by the method of simulated moments.…”

“…Alternatively, a multicriteria objective function could be considered(Kalaba and Tesfatsion 1996). Given the resulting increase in computational cost, it is left for future research to follow this approach.…”

Rate Parameter Distributions for Isobutane Dehydrogenation and Isobutene Dimerization and Desorption over HZSM-5

“…As μ approaches infinity the parameter determination approach the zero dynamic cost (ordinary least squares) solution [3,33]. When μ is set close to zero, priority is given to the dynamic specification and to estimates of the time-varying parameters [34], and hence a closer fit to the data (μ = 0.001 was used).…”