Research on the Urban Construction Status of Prefecture-Level Cities

in Heilongjiang Province Based on SPSS Analysis

Shuguang Wang

and Haoyan Wang

School of Finance and Public Administration, Harbin University of Commerce, No. 1 Xuehai Street, Harbin, China

Keywords:

SPSS Analysis, Software Cluster, Analysis Analysis of Factors, Construction of City.

Abstract:

In the different analysis of regional construction and development, many experts and scholars believe that

economic development is the main object of analysis, and use GDP index to evaluate the construction situation

(Ning, 2001). Although the data are easy to obtain and can be used for vertical comparison of regional

economic differences, the GDP indicators cannot fully and completely reflect the socio-economic

development level of a region. Therefore, the multivariate statistical analysis method can be used to analyze

the construction status of Heilongjiang Province (Wang and Zhao, 2018), which can reflect the real degree of

its construction relatively completely and comprehensively.

1 INTRODUCTION

1.1 Cluster Analysis

Cluster analysis is a multivariate statistical analysis

method to study individual classification according to

the characteristics of things. The basic principle: the

individuals in the same class have great similarities,

and individuals in different classes differ greatly;

There is a certain degree of similarity between the

study variables. According to multiple observation

indicators of the samples, the statistics that can

measure the degree of similarity between the samples

or variables are specifically found (Chen, 2015).

Based on these statistics, some variables or samples

with a greater degree of similarity are divided into

two categories. The basic principle of this method is

to directly compare the properties of all things in the

sample, group those with similar properties into one

category, and divide those with large property

differences into different categories. That is, the

difference in nature between the same kind of things

is small, and the difference in nature between classes

is large. Distance is the most commonly used way to

describe the degree of kinship between samples,

among which Euclidian distance is the most widely

used in cluster analysis. Its expression is as follows

https://orcid.org/0000-0001-6137-6915

https://orcid.org/0000-0003-4023-4453

(Tian and Zhai, 2018):





x



−x







.



Where, X

represents the observed value of the K

th index of the ith sample, X

represents the observed

value of the K th index of the J th sample, and d is the

Euclidean distance between the ith sample and the J

th sample. If d is smaller, the properties of the two

samples between i and j will be closer and closer.

Samples with similar properties can be grouped

together (He, 2015).

1.2 Analysis of Factors

Factor analysis is a data simplification technology. It

explores the basic structure of observed data by

studying the internal dependencies among many

variables, and uses a few independent unobservable

variables to represent its basic data structure.

The original variable is the explicit observable

variable, while the imaginary variable is the

unobservable latent variable and is called a factor. In

order to analyze practical problems comprehensively

and objectively, it is often necessary to consider the

research objects in many aspects and collect multiple

observation index data. If these indicators are

142

Wang, S. and Wang, H.

Research on the Urban Construction Status of Prefecture-Level Cities in Heilongjiang Province Based on SPSS Analysis.

DOI: 10.5220/0012071200003624

In Proceedings of the 2nd International Conference on Public Management and Big Data Analysis (PMBDA 2022), pages 142-146

ISBN: 978-989-758-658-3

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

analyzed one by one, it will undoubtedly lead to a

one-sided understanding of the research object, and it

will be difficult to reach a comprehensive and

consistent conclusion. Factor analysis is to consider

the mutual relationship between various indicators,

using the idea of dimension reduction to convert

multiple indicators into a few discrete comprehensive

indicators so as to make the research relatively simple.

The basic principle is to group variables according to

the size of correlation so that the correlation between

variables in the same group is higher and the

correlation between variables in different groups is

lower. Each group of variables represents a common

factor, and each observed variable can be expressed

as "the sum of the linear function of the common

factor and the special factor." The factor load matrix

and the common degree of analysis variables were

studied by constructing a factor analysis model. The

main common factors were selected according to the

variance contribution of the common factors (Wang

and Zhang, 2015). After the variance maximization of

the factors was orthogonal rotation, the regression

method was used to estimate the factor score, and the

proportion of variance contribution of each factor was

weighted and summarized to obtain a comprehensive

evaluation. Its expression is:

μ=

x











∑=

𝑚

𝑥







−𝜇





𝑥







−𝜇



2 SPSS SOFTWARE WAS USED

FOR DATA ANALYSIS

2.1 Selection of Indicators

Taking prefecture-level administrative units as

regional analysis objects, the construction status of

twelve prefecture-level cities in Heilongjiang

Province was analyzed by multivariate statistical

analysis method. The selection of indicators mainly

follows the principles of representativeness,

comprehensiveness, systematization, and

accessibility, and fourteen indicators are selected as

the basis for analysis. Its indicators are as follows(Wu

and Li, 2018): Total population at the end of X1

(10,000), X2 local fiscal revenue (100 million yuan),

X3 residential area (10,000 square meters), X4 health

institutions (number), X5 per capita green park area

(square kilometers), X6 urban water penetration rate

(%), X7 urban gas penetration rate (%), X8 highways

(kilometers), X9 number of middle schools (number),

X10 social consumption Total retail sales of defective

goods (RMB 100 million), the total output value of

X11 (RMB 100 million), the total output value of

agriculture, forestry, animal husbandry and fishery

(RMB 100 million), number of industrial enterprises

(units), urban construction area (square kilometers)

of X14. See Table 1 for details.

Table 1: Basic (original) data of twelve prefecture-level cities in Heilongjiang Province.

Number region X1 X2 X3 X4 X5 X6

1 Harbin 1000.1 339.5674 5490.2888 4393 51.53 100

2 Qiqihar 403.7 74.5843 1095.2536 2735 9.71 100

3 Mudan Jiang 227.9 54.5433 1538.7163 2238 7.70 93.4

4 Jiamusi 214.9 46.5923 447.3445 1977 8.62 92

5 Daqing 278.1 152.8677 653.4401 1431 21.16 98.8

6 Yichun 87.3 15.2430 117.7372 684 16.52 89.7

7 Qitaihe 68.5 19.2536 112.2320 634 6.10 82.5

8 Jixi 149.4 33.6601 612.6107 1064 7.43 97.7

9 Heihe 127.7 42.5270 311.6536 1001 2.19 94.6

10 Suihua 371.7 67.4869 684.2342 2194 3.76 100

11 Shuangyashan 120.3 25.4147 141.4018 1098 6.32 93.5

12 Hegang 88.7 22.9843 41.5041 705 7.54 79.6

X7 X8 X9 X10 X11 X12 X13 X14

100 877.2 450 770.9157 5183.8 1168.6479 1196 473.0

99.1 600.4 247 93.5006 1200.4 752.7777 351 131.0

Research on the Urban Construction Status of Prefecture-Level Cities in Heilongjiang Province Based on SPSS Analysis

143

80 447.9 114 382.6405 831.7 360.4715 306 92.7

90.2 605.4 127 112.5432 811.8 701.1178 319 188.0

92.5 249.6 148 529.2844 2301.1 502.3565 490 327.2

95.5 132.8 53 8.4937 295.2 202.2127 63 121.9

91 98.7 48 21.0838 206.4 72.7146 102 67.6

92.6 361.7 96 51.5093 572.4 398.2630 208 80.4

91.9 526.4 88 30.0252 614.4 524.7133 115 27.9

97.9 437.6 250 79.0378 1150.2 1058.4713 369 92.8

92.1 163.7 78 18.2568 493.9 375.1901 152 118.0

87.2 10.3 50 22.7219 340.2 210.5500 139 85.0

2.2 Cluster Analysis Was Conducted

Based on SPSS Software

Firstly, SPSS software was used for systematic

cluster analysis of the data. The analysis process and

results are as follows.

Table 2: Case Summary.

Effect

ivit

Defici

enc

Aggre

gate

Numb

er of

cases

Percen

tage

Numb

er of

cases

Percen

tage

Numb

er of

cases

Percen

tage

12 100.0 0 0.0 12 100.0

It can be seen from Table 2 that there are no lost

or unparticipated samples in the clustering process of

the selected twelve prefecture-level cities in

Heilongjiang Province, which also indicates that the

cluster analysis has carried out similar clustering on

various indicators of the twelve samples, so the next

step of analysis can be carried out.

Table 3: Clustering of samples.

Case Two clusters Three clusters

1: Harbin 1 1

2: Qiqihar 2 2

3: MudanJiang 2 3

4: Jiamusi 2 2

5: Daqing 2 2

6: Yichun 2 3

7: Qitaihe 2 3

8: Jixi 2 3

9: Heihe 2 3

10: Suihua 2 2

11:

Shuan

ashan

2 3

12: Hegang 2 3

According to the sample classification table in

Table 3, we can see the sample clustering situation

when the twelve samples are divided into two clusters

and three clusters respectively. When the samples are

clustered into three categories, the clustering results

shown in Table 4 can be obtained.

Table 4: Result of clustering.

Region Categor

Harbin 1

Qiqihar, Jiamusi, Daqing, Suihua 2

MudanJiang, Yichun, Qitaihe, Jixi,

Heihe, Shuangyashan, Hegang

Next, we use K-mean clustering method to

analyze the samples: because there is no change or

only a small change in the clustering center,

convergence is realized. The maximum absolute

coordinate change of any center is 0.000. The current

iteration is two. The minimum distance between the

initial centers is 6775.821. Table 5 can be obtained:

Table 5: Cluster member table.

The serial

numbe

Region Cluster Distance

1 Harbin 1 0.000

2 Qiqiha

2 2018.111

3 MudanJian

3 850.618

4 Jiamusi 2 1904.606

5 Da

2 2094.661

6 Yichun 3 789.013

7 Qitaihe 3 776.024

8 Jixi 3 982.068

9 Heihe 3 802.365

10 Suihua 2 1139.625

11 Shuan

ashan 3 746.527

12 Hegang 3 580.750

According to the final K-means clustering

member Table 5, we can see that twelve samples are

PMBDA 2022 - International Conference on Public Management and Big Data Analysis

144

clustered into three categories. The first category is

Harbin, the second category is Qiqihar, Jiamusi,

Daqing, Suihua, and the third category is Mudanjiang,

Yichun, Qitaihe, Jixi, Heihe, Shuangyashan and

Hegang. It can be seen that the results obtained by the

two clustering methods are consistent.

2.3 Factor Analysis Was Conducted

Base on SPSS Software

Factor analysis method in SPSS software was used to

process the data of fourteen indicators of urban

construction in twelve prefecture-level cities of

Heilongjiang Province in 2020, and the characteristic

value, contribution rate and cumulative contribution

rate of the principal factors were obtained. See Table

Table 6: The main factor characteristic value, contribution

rate and cumulative contribution rate of economic

development level of twelve prefecture-level cities in

Heilongjiang Province (%).

Principal factor 1 2

Value of characteristic 10.733 1.543

Contribution rate 76.667 11.024

Cumulative contribution rate 76.667 87.691

It can be seen from Table 6 that the eigenvalue of

the variable correlation coefficient matrix is greater

than the two main factors of one (Tang, 2007), and

the cumulative contribution rate reaches 87.961%,

which together explains 87.961% of the total variance

of the original variable. Obviously, the information

represented by the two principal factors can fully

explain and provide the information expressed by the

original data, and only 12.039% of the information is

lost. Thus, the score of the two principal factors on

each original variable is obtained Y

(See Table

7),At the same time, to obtain a comprehensive index

that can reflect the economic development level of

prefecture-level cities(∑Y),Taking the contribution

of two main factors as the weight, the comprehensive

construction scores of twelve prefecture-level cities

in Heilongjiang Province are defined as follows:

∑Y=0.87429Y

+0.12571Y

Table 7: Ranking table of comprehensive construction scores of 12 prefecture-level cities in Heilongjiang Province.

City Y

ΣY Sort

Harbin 2.80925 -0.7965 2.36 1

Qiqihar 0.40901 1.51006 0.55 2

Suihua 0.27543 1.75749 0.46 3

Daqing 0.45881 -0.97039 0.28 4

Jiamusi -0.03308 0.35229 0.02 5

Mudanjiang -0.13117 -0.81468 -0.22 6

Jixi -0.3674 0.49756 -0.26 7

Heihe -0.44954 0.79631 -0.29 8

Shuangyashan -0.54346 0.07777 -0.47 9

Yichun -0.6483 -0.31156 -0.61 10

Qitaihe -0.8901 -0.8428 -0.88 11

Hegang -0.88946 -1.25007 -0.93 12

Generally speaking, the higher the comprehensive

score, the better the regional economic development

level; If the score is greater than zero, it means that

the development level of this region is above the

provincial average development level; otherwise, it is

below the provincial average development level.

Therefore, it is necessary to actively adjust the

development ideas to promote the rapid and

coordinated development of regional construction

(Zhang, 2011).

From the comprehensive score, the economic

development level of Harbin is obviously above the

average level of provincial economic development

(∑Y>0); Qiqihar, Daqing, Suihua, Jiamusi close to

the province's average level of economic

development; Mudanjiang, Yichun, Qitaihe, Jixi,

Heihe, Shuangyashan and Hegang are significantly

lower than the average level of economic

development of the whole province.

Research on the Urban Construction Status of Prefecture-Level Cities in Heilongjiang Province Based on SPSS Analysis

145

3 CONCLUSIONS

The results of the above four regional types are

basically consistent with the economic development

status of the twelve prefecture-level cities in

Heilongjiang Province. In the future, developed areas

should use Harbin-Dalian expressway to drive the

rapid development of the suburban economy, increase

its economic radiation radius, and effectively

promote the deep development of the second and

third industries. More developed areas should take

animal husbandry and agricultural and sideline

products processing industry as the leading industries

to develop green agriculture and take the road of

"green industry." Less developed areas should

vigorously develop a port and border economic

cooperation, develop cross-border tourism between

China and Russia, implement the strategy of

sustainable development affected by tourism trade,

accelerate the transformation of resource-based cities

whose resources are close to exhaustion, and seek

development directions according to local conditions.

Backward areas should take forestry and tourism as

the leading industries, optimize the land structure,

improve the degree of industrialization, and develop

township enterprises. At the same time, Harbin--

Daqing--Qiqihar as the first-level development axis,

Mudanjiang--Jiamusi--Shuangyashan as the second-

level development axis, and radiate to Heihe, Yichun

and other cities; We should strengthen the diffusion

of economy, technology and capital from

economically developed central cities to surrounding

cities and promote the process of regional industrial

integration.

REFERENCES

Chen Leilei. Research on K-Means Text Clustering with

Different Distance Measures [J]. Software, 2015,36 (1) :

56-61.

Heilongjiang Provincial Bureau of Statistics. Heilongjiang

Statistical Yearbook [Z]. Beijing: China Statistics Press.

He Xiaoqun. Multivariate Statistical Analysis [M]. Beijing:

China Renmin University Press, 2015.

Ning Yue-min, TANG Li-zhi. The concept and index

system of urban competitiveness [J]. Modern Urban

Research, 2001, 10(3).

Tang Gongshuang. Analysis of principal component

Analysis from factor analysis based on SPSS [J].

Statistical Education, 2007, (2).

Wu Rongqiang, Li Jinhong. Classification of anode

Pressure drop in Aluminum electrolyzer based on

Cluster analysis [J]. Journal of Software, 2018, 39(3):

166-169.

Wang Xin, Zhao Nan, et al. Using SPSS software to analyze

the occurrence rules and prevention of school bullying

incidents [J]. Software, 2018,39 (1) : 159-164.

Wang Xiuxiu, Zhang Wenyu, et al. Application of improved

Principal component Cluster Analysis Method in

Education Informatization [J]. Software, 2015,36 (7) :

10-16.

Zhai Xu, Tian Mingguang, et al. Prediction of Lead-acid

Battery SOC Based on K-means Clustering and

Gaussian Process Regression Integration [J]. Journal of

Software, 2018, 39(1): 132-137.

Zhang Shijie. City Competitiveness Evaluation of Fuyang

City based on Principal Component Analysis [J].

Journal of Anhui Radio and Television University, 2011,

(4).

PMBDA 2022 - International Conference on Public Management and Big Data Analysis

146