Intuitionistic Fuzzy ANOVA and its Application in Medical Diagnosis

D.Anuradha  D. Kalpanapriya

School of Advanced Sciences, VIT University, Tamil Nadu, India.

*Corresponding Author E-mail: anuradhadhanapal1981@gmail.com

ABSTRACT:

A heuristic statisticalmethod for testing intuitionistic fuzzy hypotheses of onefactor and two factor ANOVA model using intuitionistic fuzzy data samples is proposed. In this method, the decision rules to accept or reject the null and alternative hypotheses are given. The idea of the proposed procedure has been clarified with the help of numerical examples using R studio.

KEYWORDS: Testing hypotheses, Intuitionistic fuzzy data, ANOVA model

INTRODUCTION:

Statistical hypothesis testing is an applied statistical analysis in which inference of population parameters are found using the numerical samples of the population. The data analysts have interested to learn tests of statistical hypothesis for analysing the population parameters. In an experiment study, various treatments are applied to test subjects and the response data is gathered for analysis. A critical tool for carrying out the analysis is the Analysis of Variance(ANOVA). It enables a researcher to differentiate treatment results based on easily computed statistical quantities from the treatment outcome. ANOVA is a technique of statistics that will enable us to test the null hypothesis that more than two population means are equal against the alternative hypothesis that they are not equal by using their sample data. This statistical technique is widely used in almost all areas of research. In conventional hypothesis testing Devore considering samples are precise and the significance test leads to the binary decision. However, in real life the sample data cannot be recorded precisely always. So, imprecise data sample may be got for testing the hypothesis.Intuitionistic fuzzy set (IFS) is one of the generalizations of fuzzy set theory.

Atanassov introduced the concept of IFS. The knowledge and semantic representation of IFS become more meaningful, resourceful and applicable since it includes the degree of membership function, degree of non-membership function and the hesitation margin. Intuitionistic fuzzy set is a tool in modeling real life problems like sales analysis, marketing, financial services and also it is used in different fields of science.MikihikoKonishi et al. proposed the method of ANOVA for the fuzzy interval data by using the concept of fuzzy sets. Hypothesis testing of one factor ANOVA model for fuzzy data was proposed by Wu using the h-level set and the notions of pessimistic degree and optimistic degree by solving optimization problems. Kalpanapriya and Pandian proposed a new statistical fuzzy hypothesis testing of ANOVA model for finding the significant difference among more than two population means when the data of their samples are imprecise.

Akbari et al. discussed a nonparametric method for testing the statistical hypotheses based on intuitionistic fuzzy data. Mohsen Arefi and Seyed Mahmoud Taheri proposed a least squares regression modelwhere the available data, of both explanatory variable(s) and the response variable, as well as the parameters of the model, are assumed to be Atanassov’s intuitionistic fuzzy numbers.GholamrezaHesamian and Mohamad Ghasem Akbariproposed the statistical test based on intuitionistic fuzzy hypotheses.Alireza et al. proposed an one-way ANOVA technique in fuzzy environment in which the least squares method is employed. Nourbakhsh etal. developed one-way ANOVA for a sample having fuzzy observations. In this paper, one factor and two factor ANOVA model with imprecise data is proposed. Ranking technique is adopted for ranking the imprecise data. The intuitionistic ANOVA problem has been transformed into crisp one and solved using R studio. Numerical examples are provided to illustrate the approach.

Preliminaries:

We need the following definitions of intuitionistic fuzzy set, triangular intuitionistic fuzzy number, membership and non-membership function of an intuitionistic fuzzy set/number which can be found in Atanassov.

Definition:

Let X is a nonempty set. An intuitionistic fuzzy set A in X is given by a set of ordered triples, where  define respectively, the degree of membership and degree of non-membership of the element to the set A, which is a subset of X, and for every element.

Definition: A triangular intuitionistic fuzzy number  is denoted by  where with the following membership function  and non-membership function  is given as:

Definition:

Let  and  be any two triangular intuitionistic fuzzy numbers then the arithmetic operations as follows:

Definition:

The ranking of a triangular intuitionistic fuzzy number is defined as

Definition:

Let  and be two triangular intuitionistic fuzzy numbers. The ranking of  and  by the R(.) on E , the set of triangular intuitionistic fuzzy numbers is defined as follows:

(i)if and only if

(ii) if and only if

(iii) if and only if

(iv)

(v)

One factor ANOVA Model:

The conducting of an experiment by allotting factorswhose effects are to be experimented to homogeneous experimental units by a simple random sampling design such that every unit can receive any factor with equal chance, analyzing total variation in the results into variation due to factors and variation due to chance and then testing the significance or otherwise of treatments varies over error variation. A one-factor analysis of variance method is used when the analysis involves only one factor with more than two levels and different subjects in each of the experimental conditions. In this section, we proposed the intuitionisticone-factor ANOVA model using ranking technique.Now, the one factor ANOVA model with intuitionistic fuzzy data is illustrated by the following example using R-studio.

Example:

The department of biochemistry at virginia tech study the dietary residual zinc in the blood stream for group of rats. Five pregnant rats were randomly assigned to each diet group and each was given the diet on day 22 of pregnancy. The amount of zinc in parts per million was calculated. Since under some unexpected situations, we cannot calculate it precisely. We wish to test whether there is any significant difference in residual dietary zinc among the three diets for the following intuitionist fuzzy data.

Table 1: Intuitionistic fuzzy amount of zinc in rats

 Diet 1 2 3 1 (0.30,0.50,0.70)(0.10,0.50,0.90) (0.40,0.42,0.44)(0.38,0.42,0.46) (1.04,1.06,1.08)(1.02,1.06,1.10) 2 (0.40,0.42,0.44)(0.38,0.42,0.46) (0.38,0.40,0.42)(0.36,0.40,0.44) (0.80,0.82,0.84)(0.78,0.82,0.86) 3 (0.63,0.65,0.67)(0.61,0.65,0.69) (0.71,0.73,0.75)(0.69,0.73,0.77) (0.70,0.72,0.74)(0.68,0.72,0.76) 4 (0.45,0.47,0.49)(0.43,0.47,0.51) (0.45,0.47,0.49)(0.43,0.47,0.51) (0.70,0.72,0.74)(0.68,0.72,0.76) 5 (0.42,0.44,0.46)(0.40,0.44,0.48) (0.67,0.69,0.71)(0.65,0.69,0.73) (0.80,0.82,0.84)(0.78,0.82,0.86)

Now, using the ranking technique for the above measure, we get the following table:

Table 2 :Intuitionistic fuzzy amount of zinc in rats after ranking method

 Diet 1 2 3 1 0.50 0.42 1.06 2 0.42 0.40 0.82 3 0.65 0.73 0.72 4 0.47 0.47 0.72 5 0.44 0.69 0.82

Null hypothesis for the above data is considered as  against the alternative hypothesis, . Using R studio, we obtained the following result for table 2

Table 3: Summary

 DF Sum Sq Mean Sq F Value P-value Diet 2 0.3236 0.16178 9.33 0.000359 Residuals 12 0.02081 0.01734

From table 3 it was analyzed to reject the null hypothesis for the given intuitionistic fuzzy data since, the probability value is less than the significance level 0.05. Therefore the alternative hypothesis of the given intuitionistic fuzzy data is accepted. Thus we conclude that there is a significant difference in residual dietary zinc among the three diets.

Two factor ANOVA Model:

The conducting of an experiment on experimental units, which differ in quality with respect to one character and hence stratified with respect to such changing character into different within-homogeneous blocks and then allotting treatments, whose effects are to be experimented, to homogenized units within each block by simple random sampling design independently without repetitions and thereafter splitting the total variation in the results into blocks variation, treatments variation, residual /error variation and lastly testing the significance of these variations over error variation. In this section, we concentrate on the case of two factors of interest with respect to Intuitionistic fuzzy set such that each one of factors occurs with its own number of levels. When we interest to study the effects of two factors with respect to Intuitionistic fuzzy set, it is much more advantageous to perform a two-way analysis of variance, as opposed to two separate one-factor ANOVAs. The intuitionistic two-factor ANOVA model using ranking method is proposed.Now, the two factor ANOVA model with intuitionistic fuzzy data is illustrated by the following example using R-studio.

Example:

Twenty four sheep were used for personnel in the department of animal Science to experiment at Virginia tech to study area and aqueous ammonia treatment of wheat straw. The purpose was to improve nutritional value for male sheep. The diet treatments were controlled, urea at feeding, ammonia-treated straw, and urea-treated straw and they were separated according to relative weight.

There were four sheep in each homogeneous group (by weight) and each of them was given one of the four diets in random order. Since, under some unexpected situations, we cannot measure the weight of sheep precisely. For each of the 24 sheep, the percent dry matter digested was measured. We wish to test whether the diets are homogeneous or not for the following intuitionistic fuzzy data.

Table 4: Intuitionistic fuzzy weight of the sheep under nutritional systems

 Diet Control Urea at feeding Ammonia treated Urea treated Group by Weight 1 (30,32,34)(28,32,36) (31,35,39)(27,35,43) (47,49,51)(45,49,53) (44,46,48)(42,46,50) 2 (34,36,38)(32,36,40) (36,38,40)(34,38,42) (47,53,59)(41,53,65) (38,42,46)(34,42,50) 3 (34,36,38)(32,36,40) (33,37,41)(29,37,45) (49,52,55)(46,52,58) (41,45,49)(37,45,53) 4 (38,40,42)(36,40,44) (32,34,36)(30,34,38) (41,45,49)(37,45,53) (41,45,49)(37,45,53) 5 (32,34,36)(30,34,38) (33,37,41)(29,37,45) (45,47,49)(43,47,51) (41,43,45)(39,43,47) 6 (31,33,35)(29,33,37) (32,34,36)(30,34,38) (47,49,51)(45,49,53) (45,47,49)(43,47,51)

Now, using the ranking technique for the imprecise weight, we get the following table 5:

Table 5: Weight of the sheep under nutritional systems after ranking

 Diet Control Urea at feeding Ammonia treated Urea treated Group by Weight 1 32 35 49 46 2 36 38 53 42 3 36 37 52 45 4 40 34 45 45 5 34 37 47 43 6 33 34 49 47

Null hypothesis for columns:  against

the alternative hypothesis,  Null hypothesis for rows  against

the alternative hypothesis,

Using R studio, we obtained the following result for table 6

Table 6: Summary

 DF Sum Sq Mean Sq F-Value P-value Diet 3 844.1 281.37 42.154 0.00000151 Weight 5 17.7 3.54 0.531 0.75 Residuals 15 100.1 6.67

Since probability value is less than the significance level 0.05, the null hypothesis is rejected for the given intuitionistic fuzzy data. Therefore the alternative hypothesis of the given intuitionistic fuzzy data is accepted. Thus we conclude that the diets are not homogeneous.

CONCLUSION:

In this paper, we consider one factor and two factor ANOVA problems with intuitionistic fuzzy data. The procedure of the proposed approach is explained by using real life examples. Decision makers may use the proposed ANOVA techniques in the real life issues based intuitionistic fuzzy set in real life problem involving samples for taking appropriate decisions in an easy and effective manner.

REFERENCE:

1.     Akbari, Mohammad Ghasem and Arefi, Mohsen, Statistical nonparametric test based on the intuitionistic fuzzy data, Journal of Intelligent and Fuzzy Systems, 25, 2013, 525-534.

2.     AlirezaJiryaei, Abbas Parchami and Mashaalla Mashinchi, One-way Anova and least squares method based on fuzzy random variables, Turkish Journal of Fuzzy Systems, 4, 2013, 18-33.

3.     Atanassov. K.T, Intuitionistic fuzzy sets, VII, ITKR’s Session, 1983, Sofia.

4.     Atanassov. K.T, Intuitionistic fuzzy sets, Fuzzy sets and Systems, 1986.

5.     Atanassov.K.T. New operations defined over intuitionistic fuzzy sets, Fuzzy sets and Systems, 6, 1994, 137-142.

6.     Atanassov. K.T, Intuitionistic fuzzy sets: theory and application, 1999, Springer.

7.     Devore, J.L, Probability and Statistics for Engineers, 2008, Cengage.

8.     GholamrezaHesamian and Mohamad Ghasem Akbari, Statistical Test based on Intuitionistic Fuzzy Hypotheses, Communications in Statistics - Theory and Methods , 2016,1-11.

9.     Kalpana priya. D and P. Pandian, Fuzzy hypothesis testing of Anova model with fuzzy data International Journal of Modern Engineering Research, 2, 2012, 2951-2956.

10.   Mikihiko Konishi, Tetsuji Okuda and Kiyoji Asai, Analysis of variance based on fuzzy interval data using moment correction method”, International Journal of Innovative Computing, Information and Control, 2, 2006, 83-99.

11.   Mohsen Arefi and Seyed Mahmoud Taheri, Least-Squares Regression Based on Atanassov’s Intuitionistic Fuzzy Inputs–Outputs and Atanassov’s Intuitionistic Fuzzy Parameters, IEEE transactions on fuzzy systems, 23, 2015, 1143-1154.

12.   Nourbakhsh, M. Mashinchbi and A. Parchami, Analysis of variance based on fuzzy observations, International Journal of Systems Science, 44, 2013, .714-726.

13.   Wu. H.C., Analysis of variance for fuzzy data, International Journal of Systems Science, 38, 2007, 235–246.

14.   Zadeh. L. A, Fuzzy sets Information Control, 8, 1965, 338-353.

Received on 05.07.2017          Modified on 26.09.2017

Accepted on 01.11.2017        © RJPT All right reserved

Research J. Pharm. and Tech 2018; 11(2):653-656.

DOI: 10.5958/0974-360X.2018.00122.1