Search

Probabilistic Analysis of Optimization Algorithms


Optimization algorithms transform data about decision-options under restrictions into proposals for making such decisions. Criteria of quality for such algorithms are e. g. precision, computation time, storage space.
Classical mathematics judged about such algorithms on behalf of their worst-case-behavior. In this field of research we base our judgment on the average-case-behavior.
For that purpose one supposes that the data are randomly distributed. Starting from that point one can derive expected values for the behavior of the algorithms. This is called probabilistic analysis.
For the probabilistic analysis of the Simplex-Method applied to solve Linear Optimization Problems the author has been awarded the Lanchester-Prize 1982 for the best publication of that year.

Scientists