A Modeling Language for Global Optimization
"This book describes an exciting development in computer science. Itdescribes a system which allows non-mathematicians to solve difficultnumerical problems using interval reasoning. The advanced intervalreasoning techniques used by *Numerica* enable it to handle non-linearconstraints and optimization with many benefits. As such this book isof considerable interest to a wide audience." Mark G. Wallace, IC-Parc, William Penney Laboratory, ImperialCollege, London
Pascal Van Hentenryck is Professor in the Department of Computer Science at Brown University. He is the author or editor of several MIT Press books. Laurent Michel is Assistant Professor in the Department of Computer Science and Engineering at the University of Connecticut.
Part 1 Introduction: nonlinear programming; local methods; global methods; Numerica; outline. Part 2 A tour of Numerica: getting started; generic constraints; constants; ranges; input parameters; aggregation operators; functions; sets; unconstrained optimization; constrained optimization; local constraint solving; local unconstrained optimization; soft constraints; real constraints and uncertain data; display; accuracy. Part 3 The meaning of Numerica: interval analysis; constraint solving; unconstrained optimization; interpretation of the results. Part 4 Modelling in Numerica: what can go wrong in Numerica; improving Numerica statements. Part 5 The syntax of Numerica: overall structure; expressions; the constant section; the input section; the set section; the variable section; the function section; the body section; the display section; the pragma section; scoping rules. Part 6 The semantics of Numerica: interval arithmetic; semantics of constraint solving; semantics of unconstrained minimization; semantics of constrained minimization; non-canonical boxes. Part 7 An implementation of Numerica: overview of the algorithm; domain-specific and monotonic interval extensions; constraint solving; unconstrained optimization; constrained optimization; advanced techniques; an implementation of box consistency. Part 8 Experimental results: constraint solving; unconstrained optimization; constrained optimization; appendices.