| US 7,542,932 B2 | ||
| Systems and methods for multi-objective portfolio optimization | ||
| Kete Charles Chalermkraivuth, Niskayuna, N.Y. (US); Srinivas Bollapragada, Niskayuna, N.Y. (US); Piero Patrone Bonissone, Schenectady, N.Y. (US); Michael Craig Clark, Glen Ellyn, Ill. (US); Neil Holger White Eklund, Clifton Park, N.Y. (US); Naresh Sundaram Iyer, Clifton Park, N.Y. (US); and Rajesh Venkat Subbu, Clifton Park, N.Y. (US) | ||
| Assigned to General Electric Company, Niskayuna, N.Y. (US) | ||
| Filed on Feb. 20, 2004, as Appl. No. 10/781,780. | ||
| Prior Publication US 2005/0187844 A1, Aug. 25, 2005 | ||
| Int. Cl. G06Q 40/00 (2006.01) | ||
| U.S. Cl. 705—35 [705/7; 705/10; 705/36 R; 708/130; 708/131; 708/132; 708/134; 434/107; 434/108; 434/191; 434/211] | 39 Claims |
| 1. A method for multi-objective portfolio optimization, the method comprising the steps of:
generating an initial population of solutions of portfolio allocations using a combination of linear programming and sequential
linear programming algorithms in a portfolio configuration space using a computing device, the portfolio configuration space
having a plurality of dimensions;
generating a first interim efficient frontier in a portfolio performance space having at least three dimensions using a Pareto
Sorting Evolutionary Algorithm (PSEA);
generating a second interim efficient frontier in the portfolio performance space using a Target Objectives Genetic Algorithm
(TOGA);
concatenating the first interim efficient frontier with the second interim efficient frontier to create a third interim efficient
frontier; and
passing the third interim efficient frontier through a dominance filter to generate a final efficient frontier for use in
investment decisions.
|