The term Soft Computing (SC) encompasses many techniques which include: Fuzzy Logic (FL), NeuroComputing (NC), Probabilistic Reasoning (PR), Evolutionary Computing (EC) or Genetic Algorithms (GA), Chaotic Systems (CS), Belief Network (BN) and part of Learning Theory (LT) (Zadeh, 1965, 1994, 1995; Mellit and Kalogirou, 2008). SC techniques are different from analytical approach in that they employ computing techniques that are capable of representing imprecise, uncertain and vague concepts (Voracek, 2001a; Kulak et al., 2005; Kahraman, 2007; Guarino et al., 2009). Analytical or in other words hard computing, approaches on the other hand use binary logic, crisp classification and deterministic reasoning. In their editorial review, (Hoffmann et al., 2005) observed that: “In contrast with hard computing methods that only deal with precision, certainty, and rigor, soft computing is effective in acquiring vague or sub-optimal but efficient and competitive solutions. It takes advantage of intuition, which implies the human mind-based intuitive and subjective thinking is implemented here”. Techniques in SC are able to handle non-linearity and they also offer computational simplicity when compared with the analytical methods. These techniques have been shown to be able to manage large amount of information and mimic biological systems in learning, linguistic conceptualization, optimization and generalization abilities. Soft computing techniques are finding growing acceptance in materials engineering and three of them are popular, namely: (i) Fuzzy Logic (FL), (ii) Artificial Neural Networks (ANN) and (iii) Genetic Algorithms (GA). There are well established methodologies for integrating SC techniques to realize synergistic or hybrid models with which better results could be obtained (Zadeh, 2001). The use of hybrid techniques is also growing. Real world problems have to deal with systems which are non-linear, time-varying in nature with uncertainty and high complexity. The computing of such systems is study of algorithmic processes which describe and transform information: their theory, analysis, design, efficiency, implementation, and application. Conventional computing/Hard computing requires exact mathematical model and lot of computation time. For such problems, methods which are computationally intelligent, possess human like expertise and can adapt to the changing environment, can be used effectively and efficiently. Soft computing utilizes computation, reasoning and inference to diminish computational cost by exploiting tolerance for imprecision, uncertainty, partial truth and approximation. Soft Computing with its roots in fuzzy logic, artificial neural network, and evolutionary computation has become one of the most important research field applied to numerous engineering areas such as Aircraft, Communication networks, computer science, power systems and control applications. Soft Computing Techniques comprises of core methodologies: Fuzzy Systems (FS), including Fuzzy Logic (FL); Evolutionary Computation (EC), including Genetic Algorithms (GAs); Artificial Neural Networks (ANN), including Neural Computing (NC); Machine Learning (ML); and Probabilistic Reasoning (PR). Where PR and FL systems are based on knowledge-driven reasoning, whereas, ANN and EC, are data-driven search and optimization approaches.
