TWO EXTENSIONS OF STATISTICAL DECISION THEORY
DUNLAP AND ASSOCIATES EAST INC NORWALK CT
Pagination or Media Count:
The objective of the project was to develop broader formulations of the mathematical statistical theory of decisions. This final report presents two broad scope generalizations which have resulted from the project. The first generalization discussed is a decision-making model which applies to the case of a not-well-informed decision maker with independent data sources. In this model, the inference about the prior distribution is determined from the solution of an adjunct decision problem, which specifies the minimum risk hypothesis in the light of the available information. The second generalization presented is a model of multi-period decision making for both stationary and Markovian environments. In contrast to the model discussed in the above paragraph, this model does not assume independent data sources, i.e., that the observation processes are not affected by the actions of the decision maker.
- Statistics and Probability