Research on the Forecast of Jiangxi's Digital Economy Scale

Based on EWM and BP-Neural Network Models

Linbo Chen

, Zhining Zhang

and Yu Zhao

East China University of Technology, Nanchang 330000, Jiangxi, China

Keywords: Jiangxi's Digital Economy, Economic Scale Forecast, EWM Model, BP Neural Network Model.

Abstract: Accurately judging the future development trend of the digital economy is the premise of exploring the

realization path of strengthening the digital economy of Jiangxi. We use EWM model to measure the digital

economy development index of Jiangxi Province from 2011 to 2020. Secondly, we build a predictive model

for the scale of Jiangxi's digital economy based on BP neural network method. Finally, we use scenario

analysis and the prediction model to simulate the development path and predict the scale of Jiangxi's digital

economy. The results show that: First, during 2011-2020, the scale of digital economy development in Jiangxi

Province has grown steadily, with the fastest month-on-month growth rate of 46.29% in 2017. Secondly, the

independent variables selected in this paper can positively affect the development of the digital economy in

Jiangxi Province. The correlation between residents' wage level and digital economy scale index is the

strongest. Third, the scale of Jiangxi's digital economy can achieve sustainable growth on the premise of

maintaining the current development trend or changing the development path during 2021-2025, and the

digital economy gains is the most significant under the development mode of path 4.

1 INTRODUCTION

The digital economy is a new economic form

following agricultural economy, industrial economy

and service economy. It is significant to the

transformation and upgrading of China's industrial

structure, coordinated employment quantity and

quality promotion, and high-quality economic

development. In January 2022, Chinese State Council

issued the "Fourteenth Five Year" Digital Economy

Development Plan. It is proposed in the plan that by

2025, the added value of the core industries of the

digital economy will account for 10% of GDP, the

data element market system will be initially

established, the digital transformation of the industry

will reach a new level, the level of digital

industrialization will be significantly improved, the

digital public services will be more inclusive and

equal, and the digital economy governance system

will be more perfect. To follow up the "Fourteenth

Five Year" Digital Economy Development Plan,

Jiangxi Province responded promptly, actively seized

https://orcid.org/0000-0001-6751-9681

https://orcid.org/0000-0002-2803-5294

https://orcid.org/0000-0003-1639-676X

the opportunities of digital economy development,

and introduced policies and measures tailored to local

conditions. In May 2022, Jiangxi Province issued the

"Fourteenth Five Year" Digital Economy

Development Plan of Jiangxi Province, which takes

the deep integration of digital technology and real

economy as the main line, deeply promotes the "No. 1

Development Project" to make the digital economy

better and stronger, focuses on the essential tracks

such as electronic information industry, focuses on

cultivating new tracks for VR, Internet of Things,

UAV, and other industries, and promotes the digital

transformation of traditional industries such as

agriculture, manufacturing, logistics. We will strive to

build a new highland for developing the national

digital economy and a digital industry development

cluster in the central region to achieve high-quality

leapfrog development.

The methods for measuring the scale of the digital

economy mainly include direct measurement

algorithm, indicator system method and satellite

account method (Wang S 2021). The first category

562

Chen, L., Zhang, Z. and Zhao, Y.

Research on the Forecast of Jiangxi’s Digital Economy Scale Based on EWM and BP-Neural Network Models.

DOI: 10.5220/0012036900003620

In Proceedings of the 4th International Conference on Economic Management and Model Engineering (ICEMME 2022), pages 562-569

ISBN: 978-989-758-636-1

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

uses the direct measurement method to measure the

scale of the digital economy. The US Department of

Commerce first released a report measuring the digital

economy in 1998 to analyze the impact of IT and e-

commerce development on the US economy. China

first measured the total amount of digital economy

based on growth accounting in 2008 (Kang T.x 2008).

The second category uses the indicator system method

to measure the scale of the digital economy. From the

provincial level, Liu Jun et al. (2020) selected

indicators according to the three dimensions of

information technology development, Internet

development, and digital transaction development to

establish a quality evaluation system for China's

digital economy development. Zhao Tao et al. (2020)

measured the comprehensive development level of the

digital economy from the aspects of Internet

development and digital financial inclusion from the

perspective of city-level data. The third category uses

the satellite account method to measure the scale of

the digital economy. Yang Zhongshan and Zhang

Meihui (2019), referring to the research progress of

the International Digital Economy Satellite Account

(DESA), set out from the broad concept of digital

economy and take the characteristics of digital

economic activities as the core to build a DESA that

conforms to China's national conditions. Xiang

Shujian and Wu Wenjun (2019) believed that digital

economic satellite accounts should be added based on

the existing central system, and a digital economic

accounting framework including production

accounting, income distribution accounting and

accumulation accounting was built.

In terms of forecasting the scale of the digital

economy, the relevant research in China is still

lagging, and there is less literature involved. In

forecasting the development of China's digital

economy, Li Dong et al. (2022) used principal

component analysis, multiple linear regression,

balance optimizer and other methods to build a

forecasting model by analyzing the data from 2013 to

2019. Xian Zude and Wang Tianqi (2022) observed

the time series data of the core industries of the digital

economy and found that the growth of the core

industries of the digital economy was rapid. They

chose the index model to predict the subsequent

development trend. Li Yingjie and Han Ping (2022)

measured the development trend of China's digital

economy from 2010 to 2018 based on the entropy

method and established a gray prediction model to

predict the development trend of the digital economy

from 2019 to 2028. In terms of regional digital

economy development prediction, Ji Xiaoyan (2020)

chose to use the gray Markov model to predict the

comprehensive index of digital economy development

in Zhejiang Province from 2019 to 2023, and believed

that the overall development of digital economy in

Zhejiang Province was growing, but it could be

improved better.

To sum up, scholars at home and abroad have

made beneficial explorations on the measurement

methods of the scale of the digital economy, but few

have been involved in the prediction of the scale of the

digital economy and the research on the prediction of

the development level of the regional digital economy

is even less. Therefore, this paper takes Jiangxi

Province as the research object and combines the

entropy method to construct a comprehensive

development index of digital economy from the

Internet development and digital financial inclusion to

measure the development level of digital economy in

Jiangxi Province. Based on the original data,

combined with Holt's linear trend model and fixed

annual growth rate, the BP neural network is used to

predict the development trend of Jiangxi's digital

economy in the future for a while, and six

development paths are designed to explore the optimal

path of Jiangxi's digital economy development, to

enrich the research on China's digital economy

prediction and provide decision-making basis for the

vigorous development of Jiangxi's digital economy.

2 METHODS AND MATERIALS

2.1 BP Neural Network Model

BP neural network is a method different from the

traditional regression model, characterized by

exploring the correlation between data through

parameters and activation functions. In the

application process of neural network, three types of

information processing neurons are mainly involved:

input neurons, output neurons, and hidden neurons.

The specific structure is shown in Figure 1.

Figure. 1: Structure of BP neural network.

Research on the Forecast of Jiangxi’s Digital Economy Scale Based on EWM and BP-Neural Network Models

563

The mathematical language description is as

follows: let X=(X

,...,X

) be the training set,

xi=(x

,...,x

) be the input value,

yi=(y

,...,y

) be the prediction value ,

zi=(z

,...,z

) be the real value. Denote the

threshold and weight from the input layer to the

hidden layer as b

(1)

and v

respectively, the threshold

and weight from the hidden layer to the output layer

as b

(2)

and v

respectively, and the expected accuracy

of the model as a. When the input data reaches the

threshold, the outputs of the hidden layer and the

output layer are:

() ( * )

kj rkir

zfI f vx



(1)

() ( * )

tt jqij

zgI g vz



(2)

…where I

is the input of the hidden layer, I

is the

input of the output layer, Z

is the output of the hidden

layer, Z

is the output of the output layer, f and g are

the activation functions. Then, the mean square error

E is:

()

it it

Eyz

=−



(3)

If E>a, the model will carry out step-by-step error

backpropagation. The model iteration process is as

follows:

(1) ()

jp jp

vc vc

∂

+= −

(4)

…where η is the learning rate and c is the number

of iterations. The iterative formula from the output

layer to the hidden layer is the same as (4).

2.2 Digital Economy Measurement

Framework

Based on the connotation of digital economy and the

availability of data, this paper uses the research

methods of Zhao Tao and other scholars (Zhao T.

2020, Min L.l. et al. 2022) to measure the

comprehensive development level of digital economy

by selecting indicators from two aspects of Internet

development and inclusive development of digital

finance, and determines the weight of each indicator

through entropy weight method to design a digital

economy index measurement system as shown in

Table 1 to measure the current level of digital

economy development in Jiangxi Province. For the

measurement of Internet development, Huang

Qunhui 's method (2019) was used for reference, and

four indicators were selected: Internet penetration

rate, number of Internet-related practitioners,

Internet-related output and number of mobile Internet

users. To measure inclusive development of digital

finance, refer to the method of Guo Feng et al. (2020),

and select China's digital inclusive financial index to

measure inclusive development of digital finance.

The index is jointly compiled by the Digital Finance

Research Center of Peking University and Ant Group.

Table 1: Digital Economic Index Measurement System.

Level I indicators Level II indicators Level III indicators Indicator attribute

Comprehensive

development index of

digital economy

Internet penetration

(IP)

Number of Internet users of

100 people

Number of Internet-related

practitioners

(IRP)

Proportion of computer service

and software practitioners

Internet-related output

(IRO)

Total telecom services per

capita

Number of mobile Internet

users

(MIU)

Number of mobile phone users

with 100 people

Inclusive development of

digital finance

(DF)

China Digital Inclusive

Financial Index

ICEMME 2022 - The International Conference on Economic Management and Model Engineering

564

2.3 Data Description

This paper conducts research on the scale of digital

economy development in Jiangxi Province. The time

period is selected as 2011-2020. The data on

influencing factors and digital economy index

measurement indicators are mainly from China Urban

Statistical Yearbook, China Statistical Yearbook,

Jiangxi Statistical Yearbook, Beijing University

Digital Finance Research Center and the database of

Prospective Industry Research Institute.

3 MODEL CONSTRUCTION AND

EMPIRICAL ANALYSIS

3.1 Evaluation of Digital Economy

Development Scale

Table 2: Shows the scale evaluation model of digital

economy based on entropy weight method:

Table 2: Data Similarity Measurement Table of Various Indicators of Digital Economy.

Year

Metrics (weight) Composite value

(0.19)

IRP (0.15)

IRO

(0.1)

(0.37)

(0.19)

Comprehensive development

index of digital economy

(DES)

2011 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

2012 0.0498 0.6877 0.0085 0.1252 0.2001 0.1989

2013 0.0820 0.7532 0.0268 0.2388 0.3744 0.2917

2014 0.1020 0.8999 0.0439 0.3012 0.4695 0.3605

2015 0.3351 0.6010 0.0917 0.3412 0.5745 0.3995

2016 0.4276 0.4844 0.0411 0.3893 0.6241 0.4219

2017 0.5716 1.0000 0.1245 0.5376 0.7638 0.6172

2018 0.8448 0.0516 0.4134 0.8324 0.8572 0.6802

2019 0.9493 0.4988 0.7877 0.8821 0.9309 0.8370

2020 1.0000 0.6592 1.0000 1.0000 1.0000 0.9483

It can be seen from Table 2 that the development

of digital economy in Jiangxi Province is on the rise

steadily. The development level of the digital

economy (standardization coefficient) is from 0 in

2011 to 0.9483 in 2020. Among them, the fastest

growth rate was in 2017, with a year-on-year growth

of 46.29%. This is mainly because Jiangxi Provincial

Government held a plenary meeting in September

2016, and proposed to accelerate the transformation

and upgrading of industrial digitalization closely

around the strategies of "Internet plus", "Made in

China 2025", "mass entrepreneurship, innovation",

etc. The relative decline in the development speed in

2018 is the external manifestation of the connotation

of the coupled development model of the real

economy and the digital economy. In 2019, all fields

of the digital economy began to work together, and

the development of the digital economy turned

around.

Table 3: Drivers of Jiangxi's Digital Economy Scale Index.

Variable Meaning Definition

Pgdp Level of economic growth Ratio of GDP to total population

Fdi Foreign capital dependence Ratio of foreign direct investment to GDP

Gov Degree of government intervention Ratio of government expenditure to GDP

Edu Human capital level

Ratio of the number of students in secondary institutions of higher

learnin

to the total

ulation of the re

ion

Str Industrial structure level Proportion of tertiary industry in GDP

Pwage Residents' salary level Average wage of urban non private employees in the region

Research on the Forecast of Jiangxi’s Digital Economy Scale Based on EWM and BP-Neural Network Models

565

Based on the relevant research on digital economy

by Liu Jun and others (2020), this paper selects six

core factors that may affect the scale of digital

economy in Jiangxi Province: economic growth level,

dependence on foreign capital, government

intervention, human capital level, industrial structure

level, and residents' wage level as the primary

independent variables. Pearson and Spearman

correlation analysis of the above six variables and

Jiangxi's digital economy scale variables is

conducted, respectively, and Table 4 is obtained. It

can be seen that the above six independent variables

are significantly positively correlated with Jiangxi's

Digital Economy Scale (DES). The upper right of the

main diagonal is Spearman correlation coefficient,

and the lower left is Pearson correlation coefficient in

Table 4.

Table 4: Correlation coefficient

DES

Pgdp Fdi Gov Edu Str Pwage

DES 1 1.000*** 0.673** 0.818*** 0.358 1.000*** 1.000***

Pgdp 0.986*** 1 0.673** 0.818*** 0.358 1.000*** 1.000***

Fdi 0.639** 0.632* 1 0.515 0.697** 0.673** 0.673**

Gov 0.818*** 0.787*** 0.343 1 0.406 0.818*** 0.818***

Edu 0.672** 0.671** 0.948*** 0.330 1 0.358 0.358

Str 0.965*** 0.992*** 0.633** 0.777*** 0.668** 1 1.000***

Pwage 0.987*** 0.999*** 0.634** 0.800*** 0.666** 0.993*** 1

3.2 Construction of Jiangxi's Digital

conomy Scale Prediction Model

The six factors screened above are selected as the

input variables of BP neural network prediction, and

comprehensive development index of digital

economy is the output. The data from 2011 to 2016

are used as the training set, and the data from 2017 to

2020 are used as the verification set and test set,

respectively, to build BP neural network machine

learning samples; Levenberg-Marquardt algorithm

method is selected as the training method, and the

maximum training times are 10000. The model

structure is shown in Figure 2:

Figure 2: Structure of BP neural network model

The trained BP neural network is used to predict

the validation and test samples, and the mean square

error is 0.064. The results show that the model has an

excellent fitting effect and a small mean square error

(MSE) in forecasting the scale of Jiangxi's digital

economy.

3.3 The Design of Digital Economy

Development Path in Jiangxi

Province

By changing the change rate of six influencing factors

that affect the development scale of Jiangxi's digital

economy one by one, this paper explores the growing

trend of Jiangxi's digital economy development scale

in the future under different paths:

Present status 1: Estimate the value of the

independent variable from 2021-2025 based on the

Holt linear trend model. Keep the development trend

of Jiangxi's digital economy from 2021-2025 the

same as 2011-2020, and predict the scale of Jiangxi's

digital economy from 2021-2025.

Present status 2: calculate the average growth rate

from 2011 to 2020, use this value as the fixed growth

rate to calculate various factor values from 2021 to

2025, and predict the scale of the digital economy in

the next five years.

Path 1: The level of economic growth will affect

the development of the digital economy. Path 1 is set

as follows: from 2021-2025, the average annual GDP

per capita growth rate in Jiangxi Province will be

12.18%, and other factors will maintain the current

development level.

Path 2: Attracting foreign direct investment (FDI)

can promote the development of digital economy.

Path 2 is designed as follows: from 2021-2025, the

ratio of foreign direct investment (FDI) to GDP of

Jiangxi Province, that is, the average annual growth

ICEMME 2022 - The International Conference on Economic Management and Model Engineering

566

rate of foreign investment dependency, is set to

14.88%, and other factors remain unchanged.

Path 3: Government regulation and support is an

essential guarantee for improving digital

infrastructure. Path 3 is designed as follows: from

2021-2025, the ratio of government fiscal expenditure

to regional GDP, that is, the average annual growth

rate of government intervention, is set to 22.30%, and

other factors remain unchanged.

Path 4: The development of digital economy

urgently needs a high-level labor force. Path 4 is

designed as follows: from 2021-2025, the proportion

of students in secondary and higher education

institutions to the total population of Jiangxi

Province, that is, the average annual growth rate of

human capital level is set to 18.23%, and other factors

remain unchanged.

Path 5: The level of industrial structure will play

a positive role in developing the digital economy.

Path 5 is designed as follows: from 2021-2025, the

ratio of the tertiary industry to the regional GDP, that

is, the average annual growth rate of the industrial

structure level is set to 13.53%, and other factors

remain unchanged.

Path 6: The digital economy cannot be separated

from the digital transaction behavior of residents.

Path 6 is designed as follows: from 2021-2025, the

average wage of urban non-private employers in the

region, that is, the average annual growth rate of

residents' wage level, is set to 1.35%, and other

factors remain unchanged.

Table 5 shows the parameter changes of the above

six paths:

Table 5: Parameter Changes under Each Path.

Path

Indicators (%)

Path1 Path2 Path3 Path4 Path5 Path6

Economic growth level growth rate 12.18 9.18 9.18 9.18 9.18 9.18

The growth rate of foreign capital

dependence

3.49 14.88 3.49 3.49 3.49 3.49

The growth rate of government

intervention

2.11 2.11 22.30 2.11 2.11 2.11

The growth rate of human capital level 1.95 1.95 1.95 18.23 1.95 1.95

Industrial structure level growth rate 4.13 4.13 4.13 4.13 13.53 4.13

The growth rate of household wages 10.00 10.00 10.00 10.00 10.00 1.35

3.4 Analysis of BP Neural Network

Prediction Results

Based on the above path design, the predicted value

of Jiangxi's digital economy scale from 2021-2025

obtained by BP neural network prediction model is

shown in Table 6 below:

Table 6: 2021~2025 Digital Economy Comprehensive Development Index Forecast under Each Path.

Year

Present status

Path 1 Path 2 Path 3 Path 4 Path 5 Path 6

2021 1.0881 1.0316 1.0316 0.9771 0.9808 1.0897 1.0085 1.0128

2022 1.1934 1.0739 1.1061 0.9974 1.0171 1.1859 1.0698 1.0768

2023 1.2648 1.1187 1.1640 1.0092 1.0489 1.2259 1.1249 1.1314

2024 1.3098 1.1661 1.2029 1.0144 1.0743 1.2389 1.1693 1.1726

2025 1.3334 1.2165 1.2253 1.0141 1.0936 1.2435 1.2014 1.2000

Research on the Forecast of Jiangxi’s Digital Economy Scale Based on EWM and BP-Neural Network Models

567

Figure 3 Jiangxi's Digital Economy Scale Forecast Index under Different Paths

The following conclusions can be drawn from the

above table 6 and the discount chart of predicted

value changes under different paths (Figure 3):

Present status 1: According to the independent

variable values estimated by Holt's linear trend model

from 2021-2025, if the current development trend is

maintained, the digital economy scale index of

Jiangxi Province will continue to grow between 2021

and 2025, eventually reaching 1.33.

Present status 2: The digital economy scale index

of Jiangxi Province in 2025 calculated based on the

fixed annual growth rate is 0.117 lower than that of

present status 1, which is caused by the prediction

method of impact factors.

Path 1: Improve the level of economic

development. The digital economy scale index of

Jiangxi Province will reach 1.225 by 2025, which will

exceed the digital economy scale index achieved by

maintaining the current development trend path. This

is because good performance of economic growth can

promote the healthy development of digital economy,

which is mainly reflected in the perfection of digital

infrastructure in the province, the prosperity of

digital-related industries and the level of overall

digitalization.

Path 2: Increase the dependence on foreign

capital. The digital economy scale index of Jiangxi

Province will increase year by year from 2021 to 2025

and finally reach 1.0141. Foreign investment means

additional capital investment and foreign advanced

digital technology and digital management concept,

which can accelerate the digital construction of

Jiangxi Province.

Path 3: Strengthen government intervention, and

the digital economy scale index of Jiangxi Province

will reach 1.0936 by 2025. At present, the

development level of Jiangxi's digital economy is at a

low stage, the role of market resource allocation may

fail, and government intervention can help the healthy

development of the digital economy.

Path 4: Improving the level of human capital will

enable the digital economy scale index of Jiangxi

Province to continue to rise from 2021 to 2025 and

reach 1.2435 in 2025, which is more than the index

that can be reached by the path of maintaining the

status quo. The development of the digital economy

requires a high-level talent team. High-quality labor

can promote the long-term sustainable development

of the digital economy. The software and hardware

required for digitization need high-quality labor

support.

Path 5: Improve the industrial structure. The

digital economy scale index of Jiangxi Province will

reach 1.2014 by 2025. The gradual evolution of

traditional industries from labor-intensive to capital-

intensive, technology-intensive, and knowledge-

intensive industries will have a huge impact on the

development of digital economy.

Path 6: On the premise of ensuring the steady

increase of residents' wages, Jiangxi's digital

economy scale index will reach 1.2000 by 2025.

Based on the connotation of digital economy, the

development of digital economy cannot be separated

from residents' digital trading behavior. Therefore, to

some extent, the residents' wage level controls the

development of the digital economy.

ICEMME 2022 - The International Conference on Economic Management and Model Engineering

568

4 CONCLUSIONS

This paper uses EWM method to measure the digital

economy comprehensive development index of

Jiangxi Province from 2011 to 2020 based on

building the scale measurement index system of

digital economy. Next, this paper sets multiple

development paths combined with the driving factors

of digital economy development. It uses digital

economy scale prediction model of Jiangxi Province

to predict the future development scale of Jiangxi's

digital economy under different paths. The following

conclusions are reached: First, the level of

economic growth, dependence on foreign capital,

government intervention, the level of human capital,

the level of industrial structure and the level of

residents' wages can positively affect the

development of digital economy in Jiangxi Province,

among which the level of residents' wages has the

strongest correlation with the scale index of digital

economy. Second, from 2021-2025, the scale of

Jiangxi's digital economy can achieve sustainable

growth on the premise of maintaining the current

development trend or changing the development path.

And under the development mode of path 4, the

digital economy gains most significantly.

REFERENCES

Guo F., Wang J.y., Wang F. et al. (2020). Measuring

China’s digital financial inclusion: Index compilation

and spatial charactristics. J. Economics

(Quarterly),19(04),1401-1418.

Huang Q.h., Yu Y.z., Zhang S.l. (2019). Internet

development and productivity growth in manufacturing

industry： Internal mechanism and China experiences.

J. China Industrial Economics, (08),5-23.

Ji X.y. (2020). A study on the comprehensive evaluation of

the development of digital economy in Zhejiang

Province. D. Hangzhou Dianzi University.

Kang T.x. (2008). Digital economy and its accounting

Research. J. Statistics and Decision, (05),19-21.

Liu J, Yang Y.y., Zhang S.f. (2020). Research on the

measurement and driving factors of China’s digital

economy. J. Shanghai Economic Research, (06),81-96.

Li D., Wang S., Ren X.f. (2022). Construction of prediction

model for China’s digital economy scale. J. Statistics

and Decision,38(10),5-9.

Li Y.j., Han P. (2022). Comprehensive evaluation and

prediction on China’s digital economy development. J.

Statistics and Decision,38(02).90-94.

Min L.l., Xu Z.z (2022). Digital economy, innovation

performance and high-quality economic development

—empirical evidence from Chinese cities. J. Statistics

and Decision, 38(03),11-15.

Wang S. (2021). Research on measurement method of

digital economy scale: a literature review. J. China

Economic and Trade Guide. (01),179-181.

Xiang S.j., Wu W.j. (2019) Research on the design of

China’s digital economy satellite account framework. J.

Statistical Research,36(10),3-16.

Xian Z.d., Wang T.q. (2022). Measurement and prediction

of the scale of the core industries in China’s digital

economy. J. Statistical Research,39(01),4-14.

Yang Z.s., Zhang M.h. (2019). A digital economy satellite

account: International experiences and a conceptual

design of China’s compilation program. J. Statistical

Research,36(05),16-30.

Zhao T., Zhang Z., Liang S.k. (2020). Digital economy,

entrepreneurship, and high-quality economic

development：empirical evidence from urban China. J.

Management World,36(10),65-76.

Research on the Forecast of Jiangxi’s Digital Economy Scale Based on EWM and BP-Neural Network Models

569