Abstract
This study suggests an approach for solving combinatorial optimization problems based on an understanding of learning processes in the human brain. With the information and inspiration from cognitive sciences, we suggest an approach called simulated learning, which simulates human learning and problem solving processes. Its advantage lies in the application to highly unsimulatable problems, where only a set of past solution examples is available. There are three main steps in this method. First, solutions with good performance are selected from a set of randomly generated examples. Second, the combinatorial information from the selected examples is stored into artificial memory matrices. Third, a good solution is derived by analyzing the patterns in the matrices. The results from experiment confirm the efficiency of the method.
| Original language | English (US) |
|---|---|
| Title of host publication | Proceedings - Annual Meeting of the Decision Sciences Institute |
| Editors | Anon |
| Place of Publication | Atlanta, GA, United States |
| Publisher | Decis Sci Inst |
| Pages | 513 |
| Number of pages | 1 |
| Volume | 2 |
| State | Published - 1997 |
| Externally published | Yes |
| Event | Proceedings of the 1997 Annual Meeting of the Decision Sciences Institute. Part 1 (of 3) - San Diego, CA, USA Duration: Nov 22 1997 → Nov 25 1997 |
Other
| Other | Proceedings of the 1997 Annual Meeting of the Decision Sciences Institute. Part 1 (of 3) |
|---|---|
| City | San Diego, CA, USA |
| Period | 11/22/97 → 11/25/97 |
ASJC Scopus subject areas
- Management Information Systems
- Hardware and Architecture