A Fuzzy-Logic Approach In Decision-Making

  • Neelam Tyagi
  • Jul 22, 2020
  • Artificial Intelligence
  • Machine Learning
A Fuzzy-Logic Approach In Decision-Making title banner

Due to the hurried advancement in the latest information technologies, it is plausible to accumulate, stock, shift, and consolidate massive amounts of data at comparatively low costs. In knowledge learning and data mining, there is an inclination to center on completely data-driven approaches. 

 

However, in getting true interferences, we take into consideration past knowledge, non-numeric data, and uncertainties. 

 

In order to make a connection among traditional knowledge-driven approaches and essentially data-driven approaches, Fuzzy logic can serve as knowledge aspects, description, and reasoning that have overlooked research in fuzzy set theory at a time. (Referring you the similar blog on application/branches of Artificial Intelligence)

 

Underlying the concept of a linguistic variable is a fact which is widely unrecognized—a fact which relates to the concept of precision. Precision has two distinct meanings—precision in value and precision in meaning. The first meaning is traditional. The second meaning is not. The second meaning is rooted in fuzzy logic. -- Lotfi Zadeh, 2013

 

What is Fuzzy Logic?

 

First, a bit of history, Fuzzy Logic was introduced in the year 1965 by Lotfi Zadeh that enables the explanation of the intermediate values among the conventional interpretations in terms of True/False.

 

Its quality is mirrored through its capability to handle uncover knowledge and regulating variable systems that help for determining gigantic solutions for such problems that can’t be solved via the classical logic. And that will be done by associating the variable range from the real field [0,1] rather than of {0,1}. 

 

Serve as auxiliary principles in analyzing the Fuzzy Logic, the Fuzzy set has a remarkable impact along with the Fuzzy relations and Fuzzy rule-base.

 

“Logic”, an essence section of critical human thinking and capability. 

 

Put smoothly, Fuzzy Logic is easy to understand, it has effective modes to interpret linguistics and subjective characteristics of the real computing world. It is straightforward and fleeter to address in comparison to mapping out in every single possibility.

 

Its Specialization includes;

 

  1. It can assist us to organize words into clear and concise sentences. Therefore, fuzzy logic is a process to describe the human inclination of accurate thinking that is the generalization of classical logic. 

  2. It is acknowledged as a sort of multi-values logic obtained from the fuzzy set theory. 

  3. It concentrates on inferences via indefinite expressions and lingual articulations in order to determine the marginality enigmas.

  4. It depends on the relative association degrees and motivated by human understanding and cognition methods that are ambiguous, inaccurate, partially true, or requiring sharp boundaries. 

  5. The scientific view of the Fuzzy Logic is to explain to numerous issues of approximate representation or indefinite data for furnishing the expected means to employ information and human expertise. 

  6. Noval computing operations, which are based on fuzzy logic, can be accepted in the advancement of smart and intelligent systems for decision making, identification, pattern & speech recognition, and in optimization & control.

 

Fuzzy logic has extensively adopted in Machine Learning and Artificial Intelligence through which computers are assisted to acquire knowledge and discover how to execute tasks. Fuzzy logic comes in the class of unsupervised learning methods and is applicable to ML as it delivers multivalued answers.

 

Various Examples and Applications

 

Fuzzy Logic can be used in Multi-parameter situations, in capturing or describing expert knowledge or system, for behavioral systems, approximate reasoning, and for the non-linear control system; below table describes the different applications and examples of Fuzzy Logic;

 

S.No

Applications 

Examples

1

Apparatuses

Rice-cooker, washing machine, climate regulation, vacuum-control.

2

Aerospace

Elevation command of spacecraft & satellites, the flow of wind, and mixture management dicing vehicles.

3

Automotive 

The anti-lock braking operation, traffic control, truck engine, transmissions to enhance efficiency, shift scheduling.

4

Industrial 

The anti-sway handling of cranes, climate direction, positioning practices, coal power plant, automated adjustments to coal quality, supervisory modes, humidity control.

5

Image Processing

Supervising glaucoma, edge detection, image equalization. 

6

Video Games & FX

Extensive (Lord of the Rings), Video Game, AI games.

7

Medical 

Diagnosis, regulate arterial pressure through anesthesia.

8

Business

Decision support operations, personnel assessment alongside massive workforce.

9

Financial

Banknote alteration, fund administration, market foresight, securities exchanging.

 

Decision-Making by Fuzzy Logic

 

In the realm of science, engineering & technology, and the business world, decision-making is much crucial activity. Decision making can be conducted in terms of three modes;

  1. By exploiting a mathematical model

  2. By acquiring expertise from specialists

  3. By formulating a smart and expert system 

 

But, nonetheless, no such mathematical model can be designed for the field that would not be interpreted in entirety. 

An issue with 2nd mode, people interruption isn’t permissible throughout, and obviously affording a specialist demands high costs. 

Now come to the third mode, decision-making, employing an expert system or monitor urges huge attention amid specialists and research teams. An expert and smart system are intelligent adequately for retaliating noise, vagueness(unsharpness), and unpredictability. 

 

“Decision making evades data-driven based binary computing, it initiates systems from machine learning to self-driving cars to depicts the methods humans fight with the uncertainties and vagueness of life.”

 

For such an expert system, the fuzzy logic is generally practiced, let’s learn, how?

 

Fuzzy set theory is purposely implemented in constructing the expert and smart system and monitors because of its efficiency and relationship to human augmentation. Hence, the Fuzzy Logic-based approach is used in Decision-Making. Fuzzy decision-making is an authoritative exemplar to handle expert acquaintance while an individual is building a fuzzy-model-based system.

 

Algorithm for Fuzzy decision-making process

 

Fuzzy processing incorporates the following steps: Problem identification, Fuzzification, Fuzzy interference, and rule base, Defuzzification, Interpretation, and Verification.


 

  1. In the first step, variables, criteria, and options are recognized and the decision-making task is established. 

 

  1. Here in the second step, real variables are transformed into linguistic ones, a user chooses the variables are incorporated into the knowledge base. For each indicator, all the defined values are universal.

 

Consequently, within the universal set, fuzzy sets are determined as their numbers and names. Linguistic variables are defined on the basis of fundamental linguistic variables, i.e., absolute, very high, high, medium, low, very low, no or zero reliability. 

 

The higher the attributes of fuzzy sets, the stronger the variable is expressed in detail under the universe. The validity of the procedure is addressed by the fact that the fundamental words ”understanding is, into some range, and the same for all people”. There are also various forms of the function, in the form of Λ, π, Z, and S.


An image highlights the algorithm for the Fuzzy-Decision Making process.

Algorithm of Fuzzy Decision 


  1. The third step is to determine the performance of the system by If-Then conditions, that is conditional sentences, that approve the position of a particular variable as a set. Besides for if-then, the conditional sentence also uses mathematical operators yes, no, and or. 

 

Under this step, it is also desirable to implement neural components for optimizing existing rules. Every blend of variables, befalling in the state of if-then condition, expresses. After that, the user sets the weight of a singular rule by its subjective opinion that can be modified while optimizing. 

 

The interferences of the fuzzy system implementation, considerably rely on the true-result of the defined rules, is the language variable. 

 

  1. The fourth step is to transform the fuzzy estimation of the output variable to describe the outcome of the fuzzy computation,i.e. the method of defuzzification is necessitated to accomplish numerical values. 

 

  1. The last step of the fuzzy process is the employment of an optimal option in practice. (Source)

 

 

Conclusion

 

Fuzzy logic, including its intelligence to deal with uncertainties, has been popularly referred to several engineering and scientific tasks. Many examples of the AC controller and the speed-controller mode in a self-driving car mirror us how fuzzy logic can be utilized to systems where knowledge and noise perform spontaneously.

 

“As far as the laws of mathematics refer to reality, they are not certain. And as far as they are certain, they do not refer to reality.”—Albert Einstein

 

All in once, the Fuzzy Logic is not only a set-theory but also an alternate means to glance at the world. It depicts various practical techniques to answer many problems in different fields. Moreover, the response doesn’t eternally want intricate equations, it simply demands some basic knowledge and scarce fuzzy thinking.

0%

Comments