DYNAMIC PROGRAMMING, LEARNING, AND ADAPTIVE PROCESSES
Abstract
The intensive study in recent years of a variety of descriptive and variational processes, such as those which arise in biology, psychology, engineering, and economics, has uncovered many problems which are too complex to be solved by classical mathematical techniques. In order to describe some of the difficulties involved, the essentials of the classical approach for dealing with processes of this sort, in which there is insufficient information about the state variables, are reviewed. Some of the ways in which dynamic programming and adaptive control may be used to bridge the gap between classical and modern theories are indicated. Problems encountered in the study of adaptive processes are discussed and some directions for future research are suggested.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1964
- Accession Number
- AD0605496
Entities
People
- Richard E. Bellman
Organizations
- RAND Corporation