Soft Computing Techniques

Table of Contents:

A hierarchy of soft computing techniques is shown in Figure 3.0.

Four Factors of Soft Computing

It is clear from the previous section that progress and development is visualized in the field of soft computing. A proper definition of the idea enclosed in the technique of soft computing has not been known until now. Numerous ideas from various research into the technique was found. Further, many branches of science are yet to adopt soft computing. Four major components were developed by Dr. Zadeh. A detailed description based on the four factors is illustrated in Table 3.1.

Fuzzy Logic

The mapping of an input space to output space utilizing the process of fuzzy interference can be achieved effectively through Fuzzy Logic. Fuzzy Logic is referred to as multi-valued logic. Among the conventional evaluation the inclusion of intermediate values can be performed (Novak et al., 1999). Words which possess indefinite meaning will be easily reasoned out and quantified using Fuzzy Logic. Using this method, problems such as ambiguous, doubtful, and contradictory opinions can be solved effectively.

Precision is considered comparatively significant for Fuzzy Logic. Through defining fuzzy sets and fuzzy numbers, imprecision in output and input variables can be found. The way of representing the knowledge that suits the concept was the definition for Fuzzy Logic, but the exact definition is not known.

This technique has been utilized to solve many problems. Cox (1995) explained that instances of Fuzziness will be seen when the specified information is vague. For instance, for the word “young”, a single quantitative value is not suitable. For some people, age 27 will be young and for some other people age 20 may be considered young.

FIGURE 3.0 Hierarchy of soft computing techniques.

Why Use Fuzzy Logic?

• It is easy to understand, its concepts are very simple, and its naturalness makes it nice.
• It is flexible, it adds more functionality to any given system and is easy prevented to begin again after abrasion.
• Imprecise data can be tolerated.
• It has the ability to model nonlinear functions.
• It is based on natural language.

As shown in Figure 3.1, the creation of Fuzzy Logic is achieved by merging four main concepts which are mentioned below:

• If-Then Rules.
• Logical Operations.
• Membership Functions.
• Fuzzy Sets.

The overall fuzzy process is shown in Figure 3.2.

TABLE 3.1

Soft Computing Factors Proposed by Zadeh

Factor	Explanation
Premises	(1) Extensively, the problems that are exist in the real world are found to be indefinite and indeterminate. (2) Two major criteria necessary for improving the costs are certainty and precision.
Principles	(1) Drawbacks such as approximation, partial truth, uncertainty, and imprecision must be overcome for attaining low solution cost, robustness, and tractability.
Implications	(1) In spite of employing the FL, SVM, and NN techniques in a competitive manner, these are included in soft computing as complementary'. (2) Effectively merged specified systems, for instances is referred to as “neuro- fuzzy systems”. (3) Based on the needs of the customer, such as camcorders, photocopiers, washing machines, and air conditioners, the above-mentioned system was found to be effective and is increasing.
Unique Property	(1) With the assistance of experimental data, the learning must be attained. (2) Simplification power is already seen is a soft computing technique. Through interpolating or approximation the output can be generated from input. In cases of unseen input, the output can be generated with the help of prior-learned input.

FIGURE 3.1 Fuzzy Logic concepts.