ABSTRACT
The previous chapters of this book have demonstrated that there is a wide variety of computing techniques that can be applied to particular problems. These include knowledge-based systems, computational intelligence methods, and conventional programs. In many cases the techniques need not be exclusive alternatives to each other but can be seen as complementary tools that can be brought together within a hybrid system. All of the techniques reviewed in this book can, in principle, be mixed with other techniques. This chapter is structured around four ways in which different computational techniques can be complementary:
• Dealing with multifaceted problems. Most real-life problems are complex and have many facets, where each facet may be best suited to a different technique. Therefore, many practical systems are designed as hybrids, incorporating several specialized modules, each of which uses the most suitable tools for its speci¡c task. One way of allowing the modules to communicate is by designing the hybrid as a blackboard system, described in Section 9.2.
