References

1

R. Rafac, J. P. Schiffer, J. S. Hangst, D. H. E. Dubin, and D. J. Wales, Proc. Natl. Acad. Sci USA 88 483 (1991).

2

L. T. Wille, Nature 324 46 (1986).

3

G. S. Mageras and R. Mohan, Med. Phys. 20 639 (1993).

4

L. L. Whyte, AM. Math. Mnthly. 59 606 (1952).

5

S. Kirkpatrick, C. D. Gelatt, Jr., and M. P. Vecchi, Science 220 671 (1983).

6

N.Metropolis, A. Rosenbluth, M. Rosenbluth, A. Teller, and E. Teller, J. Chem. Phys 21 1087 (1953).

7

T. Erber and G. M. Hockney, J. Phys. A 24, L1369 (1991).

8

J. R. Edmunsen, Acta Cryst. A48, 60 (1992).

9

L. Glasser and A. G. Avery, J. Phys. A 25, 2473 (1992).

10

J. R. Edmunsen, Acta Cryst. A49, 648 (1993).

11

S. Lin, Bell Syst. Tech. J. 44 2245 (1965); S. Lin and B. W. Kernighan, Oper. Res. 21 498 (1973).

12

W. H. Press, B. P. Flannery, S. A. Teukolsky, and W. T. Vetterling, Numerical Recipes (Cambridge, New York, 1990).

  
Table i: Values obtained using the method of constrained global optimization that we believe to be minimum Coulombic energies for Thomson's problem of N unit point charges on the surface of a unit sphere.

  
Table ii: Comparison of angst values for the office assignment problem using the method of constrained global optimization (CGO) and simulated annealing with the Lin--Kernighan rearrangement (SA). N is the number of offices, which is equal to the number of occupants. The column labelled ``minimum'' is the lowest value from ten runs using different initial conditions; the column labelled ``mean'' is the average of the ten values.