Neural model identification, variable selection and model adequacy
A study has been conducted to describe a methodology for neural model identification which facilitates hypothesis for both model adequacy and variable significance. The study also describes a model selection procedure to produce consistent estimators. The devised model selection procedure's performance has been evaluated using a real problem drawn from tactical asset allocation and the results were compared using regression analysis. Findings indicated the presence of non-linear relationships in modelling the equity premium.
Publication Name: Journal of Forecasting
Medicaid and the cost of improving access to nursing home care
Elasticities of Medicaid spending with respect to rate of return, of access with respect to Medicaid spending and of access with respect to quality are derived to quantify the trade-off between raising rate of return and lowering nursing home quality. It is revealed that raising rate of return for Medicaid patients' care can improve access to nursing home care.
Publication Name: Review of Economics and Statistics
- Abstracts: Demographic transition, family structure, and income inequality. Consumer expenditures and inequality: an analysis based on decomposition of the Gini coefficient
- Abstracts: Consistent model specification tests: omitted variables and semiparametric functional forms. Achieving semiparametric efficiency bounds in left-censored duration models
- Abstracts: Voting behavior and information aggregation in elections with private information. Social distance and social decisions
- Abstracts: The demand for good government. Information, health risk beliefs, and the demand for fats and oils. The demand for commodity packages: the case of telephone custom calling features
- Abstracts: The role of information in U.S. offshore oil and gas lease auctions. On the failure of the linkage principle in multi-unit auctions