MINIMIZING CONVEX FUNCTIONS OVER A SIMPLEX
BOEING SCIENTIFIC RESEARCH LABS SEATTLE WA
Pagination or Media Count:
The authors present an iteration procedure to locate the minimum of a continuously differentiable strictly convex function over the unbounded simplex in Euclidean n-space, and prove that the procedure converges to the unique minimum. This procedure is constructed to facilitate its adaptation to machine programming. Applications of this procedure to maximum likelihood estimation in certain non-parametric cases are mentioned.
- Numerical Mathematics