A study on application of fuzzy methods in entrepreneurship domain

Entrepreneurial culture is receiving a greater amount of attention by academician and practitioners. Various fields of studies on entrepreneurship domain have been analyzed using fuzzy methods for prediction. The fuzzy method’s application is believed could be utilized to obtain meaningful knowledge on the various areas of entrepreneurship domain of studies


Introduction
*Research in entrepreneurship is perceived importance for developing countries like Malaysia to boost economic progress and social adjustment. Attitudes towards opportunity for entrepreneurial activity have effects on their intention to create a new venture. The entrepreneurial intention is considered as a state of an individual mind directing and guiding them towards the development and implementation of new business concepts (Bird, 1988). The fuzzy methods approach can be used to evaluate the entrepreneurship tendencies in any organization (Hornaday, 1992) thru:  Assisting organizations improve their culture by explaining the elements of entrepreneurship by encouraging entrepreneurial activities when appropriate.  Advising on entrepreneurial activities where political structure should provide a climate in which economic innovation, organization creation and profit seeking on the market can take place.
This paper attempted to explore on Fuzzy methods applied in entrepreneurship domains. Fuzzy methods are designed to handle imprecise and complex problems. The cognitive framework of Fuzzy methods could be exploited by formalizing the way a human being interprets on the problems and situations. The integration of Fuzzy methods could be a reliable methodology for managers,

Fuzzy theory
The Fuzzy method was introduced by Zadeh (1965). Fuzzy methods are a computational methods based on human thinking. The significant concept in Fuzzy methods is the application of linguistic variables in which the variables values in the form of words or sentences in Natural Language (Zadeh, 1975). A wide particular application have found that Fuzzy Controllers and Fuzzy Reasoning approach are efficient in designing certain complex industrial and management systems, which cannot be modeled precisely under various assumptions and approximations (Tzafestas et al., 1994). Fuzzy methods can be roughly classified into five major areas (Wang, 1999): Mathematics-classical mathematical concepts are extended by replacing classical sets with fuzzy sets. b. Fuzzy logic and Artificial Intelligenceapproximations to classical logic are introduced and expert systems are developed based on fuzzy information and approximate reasoning. c. Fuzzy systems-fuzzy control and fuzzy approaches in signal processing and communications. d. Uncertainty and information-different kinds of uncertainties are analyzed. e. Fuzzy decision making-considers optimization problems with soft constraints. Fig. 1 illustrates in detail the area of Fuzzy Methods. Fuzzy methods also provide a simple way to arrive at a definite conclusion based upon vague, ambiguous, imprecise, noisy or missing input information. The prediction using Fuzzy methods could be organized in the following stages. The stages (Kaur and Aggrarwal, 2013) are:  Define the objectives-identify the parameter to control, identify the action to control the system, identify the possible response, and identify the probability of system failure modes.  Identify input and output-identify the input and output relationship. Choose a minimum variable for input to fuzzy engine.  Create rule-using the rule based structure of FL, break the problem down that escalated into a set of rules.  Fuzzy membership function-create fuzzy membership functions that define the input or output used in the rules.  Fuzzy Functions-create necessary fuzzy functions.  Results Evaluation-test the system, evaluates the results, tune the rules and membership functions, and retest until satisfactory results are obtained.

Fuzzy inference system
Fuzzy inference system is an application of Fuzzy Logic and Fuzzy Set Theory (Zadeh, 1965), which can be helpful to achieve classification tasks, offline process, simulation and diagnosis, online decision support tools and process control. FIS was adopted in several studies as a prediction model. This method was useful when the data sample includes linguistic variables or the data was from non-numerical sources such as questionnaires (Kusan et al., 2010).
The structure of FIS as shown in Fig. 2 consists of:  Knowledge Base-consists of database and rule base.
Rule base containing a number Fuzzy IF-THEN rules. A database defines Fuzzy Membership function of the fuzzy sets used in the fuzzy rules.  Process under control-perform the inference operation of the rules.  Fuzzification interface-transform inference results into crisp output.  Defuzzification interface-transform inference fuzzy results into crisp output.

Fuzzy membership function
Fuzzy Membership functions can be determined with two approaches. The first approach is to use the knowledge of human experts and the second approach is to use data collected from various sensors. There are several membership functions such as triangular, normal distribution, trapezoidal, quadratic, Gaussian (exponential) and special function (cos-function) (Reznik 1997;Wang 1999;Zhang and Liu, 2006). The shape of membership functions usually dependent on the system being studied or the application problems (Reznik, 1997). Fuzzy membership approaches are listed in detail in Table 1.

Fuzzy control and choice of parameters
Fuzzy Controller has three types which are Simple Fuzzy Controllers, Complex and/or multilevel fuzzy controllers and Adaptive and/or self-organizing fuzzy controller (Reznik, 1997). Fuzzy controllers can be easily modified and be employed with multiple inputs and outputs. The choice of fuzzy controllers is dependent on the choice of parameters. Therefore, to choose a parameter for certain conditions or problems, certain procedure needs to be followed. Table 2 presented the flow that must be pursued in order to produce a prediction using Fuzzy methods.

Application of fuzzy methods in entrepreneurship domain
There has been a significant increase in entrepreneurship domain studies using fuzzy methods. This has in turn increased academicians and practitioner's interest in various facets of entrepreneurial activities. In order to promote entrepreneurship, identifying and overcome the obstacles in every possible area of entrepreneurship domain are very important (Alroaia et al., 2012a).
Fuzzy methods have been identified beneficial to produce a certain prediction.

Fig. 2:
The structure of FIS (Wang, 1999)  The fuzzy methods in most of the studies of human behavior were used through a questionnaire. However, uncertainty in the related data leads to the notion of imprecision (Kushwaha and Kumar, 2009). These studies in entrepreneurship domain believe that fuzzy methods have the advantage to reduce uncertainty and clarity in results. Table 3 illustrates several studies that have applied fuzzy methods to extract and analyze particular information of interest in entrepreneurship domain. The flow structure to produce fuzzy prediction (Wang, 1999)  t norm and s norm calculation choice t norm-min or product operators s norm-max or algebraic sum

The implementation of fuzzy methods in entrepreneurship domain studies
The data collected in most of these studies were through questionnaires or surveys. Questionnaires are usually designed to assess many domains on issues related to psychology such as perceptions, opinions, emotional states, etc. The questionnaires responses usually distributed using Likert scales (Suárez et al., 2013;Castillo et al., 2014). Therefore, in order to exploit individual differences in responding questionnaires, an expressive scale should be exploited. The questionnaires using Likert scales usually require respondents to choose one within a list of prefixed labels. Analyze specific conditions of social entrepreneurs' confidence in managing their business Fuzzy Sets (Munoz and Kibler, 2016) 3 Analyze various characteristics to distinguish which entrepreneurs will sustain in their business Fuzzy Sets (Munoz, 2012) 4 Analyze the necessary and sufficient conditions for higher entrepreneurs rates Fuzzy Logic (Ferreira and Dion sio, 2016) 5 Identify business opportunity based on the factors related to entrepreneurial activities Fuzzy AHP (Sheela and Murthy, 2015) 6 Analyzing students' entrepreneurial intention based on emotional intelligence and personality traits Fuzzy DEMATEL (Dehkordi et al., 2012) 7 Analyze on the obstacles to develop entrepreneurship in the industries. Identify the critical external and internal obstacles that will hinder the development Analyze the relationship between organizational entrepreneurship and social capital to encourage people changing the organization from no entrepreneurship to entrepreneurship Fuzzy Logic (Yaghoubi et al., 2011) 12 Measure the entrepreneurial orientation to determine the degree of entrepreneurial behaviors of the firms Fuzzy AHP (Rezaei et al., 2013) 13 Evaluate the strength and gaps of the technological entrepreneurship capabilities of high tech firms Fuzzy Logic (Hejazi and Seifollahi, 2016) 14 Evaluate the priority factors in the establishment of an entrepreneurial university Fuzzy AHP (Nikfarjam et al., 2013) 15 Identify the ranking on best online business course programs conducted by a few universities Fuzzy VIKOR (Nisel, 2014) 16 Identify students entrepreneurial competencies quality Neuro Fuzzy (Arafeh, 2016) 17 Identify the influence of social capital, entrepreneurial alertness and entrepreneurship environment on business performance Fuzzy set Qualitative Comparative Analysis (Liu et al., 2016) 18 Identify the rank and the effective factors on the success of entrepreneurs which will give impact on the development in the industrial section Fuzzy DEMATEL (Alroaia et al., 2012b) However, the questionnaires based on fuzzy have a format that combines visual analogue and fuzzy linguistic scale when analyzing responses. The novelties of analyzing data using fuzzy methods that each data are treated entirely therefore relevant information will not lost (Angeles et al., 2015).The studies illustrate in Table 3 have chosen certain fuzzy methods that are possible to solve the encountered problems. The strength of these fuzzy methods was chosen because these researchers believed the studies could be solved efficiently. The strength of these fuzzy methods was further described in Table 4. Fuzzy AHP (Saaty, 1987)

AHP-Analytical Hierarchy Process
A systematic method to solve complex and multi-level decision making problems. This method is applicable in situations where decision makers and experts are available. This method able to solve hierarchical fuzzy decision making problems. (Gabus and Fontela, 1973;Gabus and Fontela, 1972)

DEMATEL-Decision Making Trial and Evaluation Laboratory
A structural model that gathers group knowledge and visualize the causal relationship of criteria by using graphical diagram. This is a decision making method in the case that several criteria have complex relationships. This method allows extraction on interdependencies and strength among the criteria.
Fuzzy DELPHI (Kaufmann and Gupta, 1988) DELPHI The method is used for structuring a group communication process to facilitate group problem solving and to structure models (Linstone et al., 1975). The method can also be used as a judgment, decision-aiding or forecasting tool (Rowe and Wright 1999), and can be applied to program planning and administration (Delbecq et al., 1975). The Delphi method can be used when there is incomplete knowledge about a problem or phenomena (Adler and Ziglio, 1996;Delbecq et al., 1975). The method can be applied to problems with subjective judgments of individuals on a collective basis (Adler and Ziglio, 1996) and focus collective human intelligence on the problem at hand (Linstone et al., 1975) , 1981) solution and the farthest distance from the negative ideal solution.
Fuzzy VIKOR (Opricovic, 1998;Opricovic and Tzeng, 2007;Opricovic and Tzeng, 2004; VIKOR-Vlse Kriterijumska Optimizacija I Kompromisno Resenje pronounce in Serbian which means Multi criteria Optimization and compromise solution This method able to solve MCDM problem with conflicting or noncommensurable criteria (Opricovic and Tzeng, 2004). A set of alternatives is ranked and selected under conflicting criteria then each alternative is evaluated according to each criterion function. The compromise rank is selected by comparing the measure of closeness to the compromise alternative (Opricovic and Tzeng, 2004;Opricovic and Tzeng, 2002). The compromise solution will be basis for negotiations which involve decision makers' preference criteria weight (Opricovic, 2009). Neuro Fuzzy (Jang, 1993) Artificial Neuro fuzzy Inference Systems (ANFIS) This method is used to achieve the reasoning and learning capabilities of Fuzzy Logic and Neural Network. Fuzzy Set QCA (Ragin, 2000;Rihoux and Ragin, 2008) Fuzzy set Quality Comparative Analysis (fsQCA) Enables to draw conclusion about logical relationships without having to reduce the data to crisp binary sets.

Conclusion
Fuzzy methods can be versatile and flexible tool for data that are complex, vague and imprecise. Fuzzy addresses application that resembles human in decision making. Fuzzy methods have the ability to generate precise solutions for certain or approximate information and the data generates through fuzzy methods has the advantage of reducing uncertainty. Liu HW, Lin YL, Xu F, and Wang H (2016