##
**Simulated annealing: theory and applications.**
*(English)*
Zbl 0643.65028

“Simulated annealing” is a computer algorithm resembling the Monte Carlo method and Southwell’s old relaxation method but more sophisticated than either. The book gives a basic description of the algorithm which can take two forms, homogeneous or inhomogeneous. The next chapter describes its relation with mathematical physics and optimization problems in general. Among further applications of the algorithm are to circuit design and image processing and the theory of neural networks.

An assessment of the performance of the algorithm, which was only invented in 1982, concludes that its performances is superior to that of other algorithms in a number of cases. A good feature of it is that it does not waste too much effort seeking an absolute optimum in cases like the travelling salesman problem where there are a vast number of near- optimum solutions differing little from the true optimum.

The book is well printed and produced and is written in an easily intelligible style. There is an extensive bibliography.

An assessment of the performance of the algorithm, which was only invented in 1982, concludes that its performances is superior to that of other algorithms in a number of cases. A good feature of it is that it does not waste too much effort seeking an absolute optimum in cases like the travelling salesman problem where there are a vast number of near- optimum solutions differing little from the true optimum.

The book is well printed and produced and is written in an easily intelligible style. There is an extensive bibliography.

Reviewer: H.N.V.Temperley

### MSC:

65K05 | Numerical mathematical programming methods |

65C05 | Monte Carlo methods |

65-02 | Research exposition (monographs, survey articles) pertaining to numerical analysis |

92Cxx | Physiological, cellular and medical topics |

94C30 | Applications of design theory to circuits and networks |

90Cxx | Mathematical programming |