Soft Computing Applications

Soft computing (SC) is a concept that was introduced by Zadeh (1992), the discoverer of fuzzy logic. He envisioned SC as being...concerned with modes of computing in which precision is traded for tractability, robustness and ease of implementation. For the most part, SC encompasses the technologies of fuzzy logic, genetic algorithms, and neural networks, and it has emerged as an effective tool for dealing with control, modeling, and decision problems in complex systems. Briefly, fuzzy logic is used to deal with imprecision and uncertainty, genetic algorithms are used for search and optimization, and neural networks are used for learning and curve fitting. In spite of these dichotomies, there are natural synergies between these technologies.

A survey of neural network applications in actuarial science is provided in this segment of the pager. For the most part, these areas of application include classification, asset and investment models, insolvency studies, and mortality and morbidity studies. Since this is a relatively new area of analysis, a number of the studies also include comparisons with rival approaches.

Components of soft computing include :

• Neural networks (NN)
• Fuzzy logics (FL)
• Evolutionary computation (EC)
• Evolutionary algorithms
• Genetic algorithms
• Differential evolution
• Metaheuristic and Swarm Intelligence
• Ant colony optimization
• Bees algorithms
• Bat algorithm
• Cuckoo search
• Harmony search
• Firefly algorithm
• Artificial immune systems
• Particle swarm optimization
• Ideas about probability including:
• Bayesian network
• Chaos theory
• Perceptron

Generally speaking, soft computing techniques resemble biological processes more closely than traditional techniques, which are largely based on formal logical systems, such assentential logic and predicate logic, or rely heavily on computer-aided numerical analysis (as in finite element analysis). Soft computing techniques are intended to complement each other.

Unlike hard computing schemes, which strive for exactness and full truth, soft computing techniques exploit the given tolerance of imprecision, partial truth, and uncertainty for a particular problem. Another common contrast comes from the observation that inductive reasoning plays a larger role in soft computing than in hard computing.