Loading...
Developing tabu search with intensification and diversification for the seriation problem
Author(s)
Date Issued
June 15, 2018
Type
Conference Paper
Abstract
The seriation problem is an important problem in combinatorial optimization. The goal of seriation is to find a linear order for data objects to reveal structural information given a loss or a merit objective function as an objective function. For the seriation problem, finding an optimal solution using exact algorithms (e.g., branch-and-bound) to find the optimal solution is currently impractical for problems with more than 35 objects. In this paper, we develop a new heuristic procedure to maximize the gradient measure which is our selected merit objective function. The proposed heuristic incorporates search intensification and diversification into standard tabu search (TS) algorithm. From our experimental results, it shows that intensification and diversification tabu search (IDTS) outperforms the standard TS algorithms in terms of efficiency, effectiveness, and robustness.
Citation
2018 5th International Conference on Industrial Engineering and Applications Iciea 2018, 279-283, 2018