Fuzzy c means_realestate_application

FUZZY CLUSTERING IN REAL ESTATE
Petr Dostál
Brno University of Technology, European Polytechnic Institute, Kunovice
Abstract
There are different methods used as supports for decision-making processes. Fuzzy
logic has had successful applications in business. In this article the fuzzy clustering
method is used in the field of real estates. The fundamental of fuzzy logic clustering is
mentioned. The case of use is presented on real estate’s clustering by means of fuzzy
logic.

1

Introduction

The soft computing plays very important roles also in real estate decision making. The
application of the fuzzy clustering is realized on the case of real estate grouping. Popular notions of
clusters include groups with low distances among the cluster members. The fuzzy clustering could be
used, not only neural networks or evolutionary algorithms. The program MATLAB® with Fuzzy
Logic Toolbox is used.

2

Theory

Cluster analysis or clustering is the task of grouping a set of objects in such a way that objects
in the same group (called cluster) are more similar (in some sense or another) to each other than to
those in other groups (clusters). In hard clustering, data is divided into distinct clusters, where each
data element belongs to exactly one cluster. In fuzzy clustering (also referred to as soft clustering),
data elements can belong to more than one cluster, and associated with each element is a set of
membership levels. These indicate the strength of the association between that data element and a
particular cluster. Fuzzy clustering is a process of assigning these membership levels, and then using
them to assign data elements to one or more clusters.
One of the most widely used fuzzy clustering algorithms is the fuzzy c-means algorithm [3].
The fuzzy c-means algorithm attempts to partition a finite collection of n elements X={x1, x2, …. , xn}
into a collection of c fuzzy clusters with respect to some given criterion. Given a finite set of data, the
algorithm returns a list of c cluster centers where each element and a partition matrix W = wij ∈ [0,1],
i=1,2, … , n, j = 1, 2, …. , c, where each element wij tells the degree to which element xi belongs to
cluster cj. the fuzzy c-means aims to minimize an objective function. The standard function is
wk(x) =

∑
j

1
d ( center
(
d ( center

k ,x
j,x

)
)

2 /( m − 1 )

,

this differs from the k-means objective function by the addition of the membership values uij and the
fuzzifier m. The fuzzifier m determines the level of cluster fuzziness. A large m results in smaller
memberships wij and hence, fuzzier clusters. In the limit m = 1, the memberships wij converge to 0 or
1, which implies a crisp partitioning. In the absence of experimentation or domain knowledge, m is
commonly set to 2. The basic fuzzy c-means algorithm, given n data points (x1, x2, …. , xn), to be
clustered, a number of c clusters with (c1, c2, …. , cn) and m the level of cluster fuzziness with.
In fuzzy clustering, every point has a degree of belonging to clusters, as in fuzzy logic, rather
than belonging completely to just one cluster. Thus, points on the edge of a cluster may be in the
cluster to a lesser degree than points in the center of cluster. An overview and comparison of different
fuzzy clustering algorithms is available.

Any point x has a set of coefficients giving the degree of being in the k-th cluster wk(x). With fuzzy
c-means, the centroid of a cluster is the mean of all points, weighted by their degree of belonging to
the cluster

∑ w
∑ w

( x )x

k

ck =

x

k

(x)

.

x

The degree of belonging, wk(x), is related inversely to the distance from x to the cluster center
as calculated on the previous pass. It also depends on a parameter m that controls how much weight is
given to the closest center.

3

Real Estate Fuzzy Clustering

The application of fuzzy clustering is realized on the cases study of real estates. The solved
clustering is based on sorting of real estate’s according their parameters. In other words, we have to
find the real estate’s with similar parameters. The variables are as follows: Price, Region and Area.
Data are represented by 46 objects. See Table 1.
Order

Region

Area

Price

Cluster

1

21

242

600

♦

2

12

1043

1650

×

3

9

113

550

♦

4

10

929

1900

5

10

446

960

6

16

511

1300

7

21

366

800

×
∗
∗
∗

8

10

151

360

♦

9

11

371

780

10

10

223

800

11

9

221

680

12

15

232

750

∗
∗
∗
∗

13

21

199

613

♦

14

21

214

620

♦

15

12

223

500

♦

16

19

232

515

♦

17

5

186

630

♦

18

21

187

501

♦

19

15

167

520

♦

20

10

260

900

∗

21

26

167

460

♦

22

13

1906

2200

×

23

28

141

425

♦

24

5

1120

1800

×

25

26

177

450

♦

26

15

164

350

♦

27

19

186

370

♦

28

14

149

288

♦

29

17

145

300

♦

30

15

121

375

♦

31

10

465

102
0

∗

32

15

120

310

♦

33

15

280

520

♦

34

10

125
0

35

13

488

188
8
116
0

36

14

394

847

37

10

233

575

♦

38

11

235

733

39

9

221

680

∗
∗

40

11

144

477

♦

41

15

164

350

♦

42

14

149

288

♦

43

16

154

434

♦

44

20

223

568

♦

45

21

203

521

♦

46

27

162

445

♦

×
∗
∗

Table 1. Real estate’s data
The output will be the classification of real estate’s according their characteristic to clusters.
The software MATLAB and its Fuzzy Logic Toolbox is used for the software applications. The
example presents the objects recorded in MS Excel format in FCr.xlsx file. This task is solved by the
program FCr.m. See Program 1.
ffd=xlsread('FCr.xlsx','data');
plot3(fd(:,1),fd(:,2), fd(:,3), 'o','color','k', 'markersize',7,'LineWidth',2)
title('Data');
xlabel('Region');ylabel('Area');zlabel('Price')
grid
[center,U,objFcn] = fcm(fd,3);
figure
plot(objFcn)
title('Fitness Function Values')
xlabel('Iteration Count')
ylabel('Fitness Function Value')
maxU = max(U);
index1 = find(U(1, :) == maxU);
figure
center
c1='x'
fd(index1,:)
c2='d'
fd(index2,:)
c3='*'

fd(index3,:)
plot3(fd(:,1),fd(:,2), fd(:,3), 'o','color','k', ...
'markersize',7)
hold on
grid
stem3(center(1,1),center(1,2),center(1,3),'marker','x','color','g','markersize',10,'LineWidth',2)
stem3(center(2,1),center(2,2),center(2,3),'marker','d','color','r','markersize',10,'LineWidth',2)
stem3(center(3,1),center(3,2),center(3,3),'marker','*','color','b','markersize',10,'LineWidth',2)
view(30,30)
line(fd(index1, 1), fd(index1,2), fd(index1,3),'linestyle','none','marker', '+','color','g');
line(fd(index2,1),fd(index2,2), fd(index2,3),'linestyle','none','marker', 'd','color','r');
line(fd(index3,1),fd(index3,2), fd(index3,3),'linestyle','none','marker', '*','color','b');
title('Real Estate Fuzzy Clustering');
xlabel('Region');ylabel('Area');zlabel('Price'))
Program 1. M-file FCr.m
The program is started using the command FCr in the MATLAB program environment. The
number of clusters is set up to 3. During the calculation the iteration count is displayed. When the
calculation is finished the output results, the coordinates of centroids and assign of product to
centroids are displayed. See Result 1.
Iteration count = 1, obj. fcn = 6473942.582957
……
center = 1.0e+03 *
0.0163 0.1668 0.4281
0.0098 1.2087 1.864
0.0126 0.3384 0.8534
c1 = ♦
ans =
21.0000 242.0000 600.0000
9.0000 113.0000 550.0000
10.0000 151.0000 360.0000
21.0000 199.0000 613.0000
21.0000 214.0000 620.0000
12.0000 223.0000 500.0000
19.0000 232.0000 515.0000
5.0000 186.0000 630.0000
21.0000 187.0000 501.0000
15.0000 167.0000 520.0000
26.0000 167.0000 460.0000
28.0000 141.0000 425.0000
26.0000 177.0000 450.0000
15.0000 164.0000 350.0000
19.0000 186.0000 370.0000
14.0000 149.0000 288.0000
17.0000 145.0000 300.0000
15.0000 121.0000 375.0000
15.0000 120.0000 310.0000
15.0000 280.0000 520.0000

10.0000
11.3333
15.0000
14.0000
15.5000
20.0000
20.5000
26.6667

232.9000
143.6667
164.0000
148.6000
154.4500
222.8500
203.3250
161.5000

575.0000
476.6667
350.0000
288.0000
433.7500
567.5000
521.0000
445.0000

c2 = ×
ans =1.0e+03 *
0.0120 1.0430
0.0100 0.9290
0.0130 1.9060
0.0050 1.1200
0.0100 1.2495

1.6500
1.9000
2.2000
1.8000
1.8875

c3 = ∗
ans =1.0e+03 *
0.0100 0.4460
0.0160 0.5110
0.0210 0.3660
0.0110 0.3710
0.0100 0.2230
0.0090 0.2210
0.0150 0.2320
0.0100 0.2600
0.0100 0.4650
0.0130 0.4878
0.0140 0.3943
0.0107 0.2353
0.0090 0.2210

0.9600
1.3000
0.8000
0.7800
0.8000
0.6800
0.7500
0.9000
1.0200
1.1600
0.8467
0.7333
0.6800
Results 1. Results of calculation

The program displays the graph where each real estate is represented by circle according its
Price, Region and Area. See Figure 1.

Figure 1. Three-dimensional graph – real estates
It is suitable to search the fitness function values dependent on number of iteration. The graph
presents good process of iteration. See Figure 2.

Figure 2. Fitness function values
The results are presented by coordinates of clusters and assignment of real estates to the
clusters. A three-dimensional stem graph is drawn. See Figure 3. See also right column Cluster of
Table 1.

Figure 3. Graph – Real estate clustering

4

Conclusion

The results are presented by centroids of three clusters marked ♦ , × , ∗ and assignment of real
estates to the clusters. The results presents the case where the cluster ♦ includes the group of real
estates of middle price, good region and middle area, the cluster ∗ includes the group of real estate’s
of high price, in excellent region and high area, the cluster × includes the real estate’s low price, bad
region and small area. The fuzzy model enables to cluster the real estate’s according their price, region
and area. The tasks from practice lead to multi-dimensional ones, where their graphical presentation is
impossible: the image of the solution is in a hyper sphere, when the variables could be price, region,
area, type of real estate, number of rooms, number of floors, equipment of real estate etc. The example
mentioned above is an application of the use of fuzzy logic for decision making of real estate’s
managers.

References
[1]
[2]
[3]
[4]
[5]
[6]
[7]

Aliev, A., Fazlollahi, B., & Aliev, R. (2004). Soft computing and its applications in business and
economics. USA: Springer.
Altroc, C. (1996). Fuzzy logic & neurofuzzy - applications in business & finance. USA: Prentice
Hall.
Bezdek, James C. (1981). Pattern Recognition with Fuzzy Objective Function Algorithms. USA:
Springer.
Chen, S., Wang, P., & Wen, T. (2004). Computational intelligence in economics and finance.
USA: Springer.
Chen, S., Wang, P., & Wen, T. (2007). Computational intelligence in economics and finance.
Volume II, USA: Springer.
Dostál, P. (2011). Advanced decision making in business and public services. Czech Republic:
CERM Academic Publishing House.
Dostál, P. (2012a). The use of soft computing for optimization in business, economics, and
finance. Meta-Heuristics Optimization Algorithms in Engineering, Business, Economics, and
Finance, USA: IGI Globe.

[8]
[9]

[10]
[11]
[12]
[13]
[14]
[15]
[16]

Dostál, P. (2012b). The use of optimization methods in business and public services. Handbook of
Optimization, USA: Springer.
Dostál, P. (2013a). The use of soft computing methods for forecasting in business, their
applications in practice. Nostradamus: Modern Methods of Prediction, Modeling and Analysis of
Nonlinear Systems, USA: Springer.
Dostál, P. (2013b). The Use of Soft Computing in Management. Handbook of Research on Novel
Soft Computing Intelligent Algorithms: Theory and Practical Applications, USA: IGI Globe.
Gil-Lafuente, A.M., Gil-Lafuente, J., & Merigó-Lindahl, J.M., (2012). Soft computing in
management and Business Economics, USA: Springer.
Ribeiro, R., & Yager, R. (1999). Soft Computing in Financial Engineering, USA: Springer.
Ruan, D., & Fedrizzi, M. (2001) Soft Computing for Risk Evaluation and Management, vol 76,
USA: Springer.
Vasant, P. (2003). Application of fuzzy linear programming in production planning, Fuzzy
Optimization and Decision Making, 2 (3), pp. 229-241.
Zadeh, L.A. (1965). Fuzzy sets. Information and Control 8, 338-353.
Zadeh, L.A. (2012). A Definition of Soft Computing - adapted from L.A. Zadeh. Retrieved January
10, 2013, from http://www.soft-computing.de/def.html.

Petr Dostál
Faculty of Business and Management, Department of Informatics, Kolejní 4, 612 00 Brno
European Polytechnic Institute, Osvobození 699, 686 04 Kunovice

Fuzzy c means_realestate_application

More Related Content

What's hot (20)

Viewers also liked (13)

Similar to Fuzzy c means_realestate_application (20)

More from Cemal Ardil (20)

Recently uploaded (20)

Fuzzy c means_realestate_application