Implementing an algorithm to solve large sequential decision analysis models

An implementation is presented of an algorithm to solve large sequential decision analysis models. This implementation in the Pascal programming language does not use special features of Pascal, and is therefore straightforward to translate into other programming languages. Empirical results show that the algorithm requires only a modest personal computer to quickly solve decision trees with several hundred thousand endpoints. This allows the use of decision analysis methods for a variety of management and engineering decisions that are often considered too complex for decision analysis. >