制造业数字化转型驱动因素研究 ——基于机器学习的证据

作者	贾映彬
姓名汉语拼音	Jia Yingbin
学号	2021000009007
培养单位	兰州财经大学
电话	17693179812
电子邮件	982332571@qq.com
入学年份	2021-9
学位类别	学术硕士
培养级别	硕士研究生
学科门类	管理学
一级学科名称	工商管理
学科方向	企业管理
学科代码	120202
第一导师姓名	关辉国
第一导师姓名汉语拼音	Guan Huiguo
第一导师单位	兰州财经大学
第一导师职称	教授
题名	制造业数字化转型驱动因素研究 ——基于机器学习的证据
英文题名	Research on drivers of digital transformation in manufacturing enterprises -- Evidence based on machine learning
关键词	企业数字化转型机器学习制造业上市公司
外文关键词	Digital transformation of enterprises ; Machine learning ; Manufacturing ; Publicly listed companies
摘要	在数字化时代，制造业企业面临着转型为数字化企业的挑战，这成为提高企业竞争力的重要途径。然而，关于制造业数字化转型的驱动因素的研究尚不够充分，现有研究多是考虑单一因素对企业数字化转型的影响。本研究旨在深入了解影响制造企业数字化转型程度的关键影响因素及其重要性，通过综合内外部因素，构建一个全面的企业数字化转型程度驱动因素研究框架。本文使用 2017 年至 2021 年上市制造业企业数据，建立了一个制造业数字化转型的预测模型，并运用多种机器学习算法，包括 Lasso 回归、支持向量机、随机森林和LightGBM，XGBoost进行建模分析。结果表明，随机森林和LightGBM，XGBoost 在拟合效果上表现最佳。然后使用 LightGBM 算法模型，通过排列重要性和 SHAP 框架下的方法，对驱动因素进行特征重要性分析，定量评估了各因素对制造企业数字化转型程度的重要性。本文发现，数字技术投资、行业竞争强度、地区数字环境、高管风险偏好、知识密集度和管理者能力在众多模型中表现出对企业数字化转型有较高的贡献度。针对这些重要因素，利用 Shap value 量化了特征对预测值影响的大小和方向（正或负），并构建了可视化工具，包括摘要图和部份依赖图，并进一步分析了它们对企业数字化转型各个维度的具体影响及不同驱动因素间的交互效应。本文的结果显示，在众多驱动因素中，数字技术投资是最重要的驱动因素。行业竞争强度和知识密集度均较高时，可能会削弱对数字化转型的正向贡献。研究还发现，高管风险偏好、地区数字环境和管理者能力也与数字化转型密切相关。本文的研究结果为制造业企业在数字化转型过程中提供了重要的决策参考，有助于企业更好地推进数字化转型，提高数字化水平，增强企业竞争力。企业应加大数字技术投资，关注行业竞争态势，合理配置知识密集度，优化高管结构，关注地区数字环境，提升管理者数字化能力，制定全面的数字化转型战略，以顺利推进数字化转型。本文的研究结论具有针对性，主要针对制造业，未来研究可扩展到其他行业，探讨不同行业数字化转型的差异性。
英文摘要	In the digital age, manufacturing enterprises are faced with the challenge of transforming into digital enterprises, which has become an important way to improve the competitiveness of enterprises. However, the research on the driving factors of the digital transformation of manufacturing industry is not enough, and most of the existing studies consider the impact of a single factor on the digital transformation of enterprises. The purpose of this study is to deeply understand the key factors affecting the degree of digital transformation of manufacturing enterprises and their importance, and to build a comprehensive research framework on the drivers of the degree of digital transformation of enterprises by integrating internal and external factors. This paper uses the data of listed manufacturing enterprises from 2017 to 2021 to establish a prediction model for the digital transformation of manufacturing industry, and uses a variety of machine learning algorithms, including Lasso regression, support vector machine, random forest and LightGBM, XGBoost for modeling analysis. The results show that Random Forest, LightGBM and XGBoost have the best fitting effect. Then, the LightGBM algorithm model is used to conduct a feature importance analysis of the driving factors through the method of ranking importance and SHAP framework, and the importance of each factor to the degree of digital transformation of manufacturing enterprises is quantitatively assessed. This paper finds that digital technology investment, industry competition intensity, regional digital environment, executive risk appetite, knowledge intensity and manager ability have a high contribution to enterprise digital transformation in many models. In view of these important factors, Shap value is used to quantify the magnitude and direction (positive or negative) of the influence of features on the predicted value, and visual tools are constructed, including summary chart and partial dependency chart, and their specific impact on various dimensions of enterprise digital transformation and the interaction effects among different drivers are further analyzed. The results of this paper show that among the many drivers, investment in digital technologies is the most important driver. Higher levels of industry competition and knowledge intensity may undermine positive contributions to digital transformation. The study also found that executive risk appetite, regional digital environments, and manager competencies are also strongly related to digital transformation. The research results of this paper provide important decision-making reference for manufacturing enterprises in the process of digital transformation, and help enterprises to better promote digital transformation, improve digital level, and enhance enterprise competitiveness. Enterprises should increase investment in digital technology, pay attention to industry competition, rationally allocate knowledge intensity, optimize executive structure, pay attention to regional digital environment, improve managers' digitalcapabilities, and formulate comprehensive digital transformation strategies to smoothly promote digital transformation. The research conclusions of this paper are targeted, mainly for the manufacturing industry, and the future research can be extended to other industries to explore the differences of digital transformation.
学位类型	硕士
答辩日期	2024-05-26
学位授予地点	甘肃省兰州市
语种	中文
论文总页数	73
参考文献总数	86
馆藏号	0006183
保密级别	公开
中图分类号	F27/220
文献类型	学位论文
条目标识符	http://ir.lzufe.edu.cn/handle/39EH0E1M/36958
专题	工商管理学院
推荐引用方式 GB/T 7714	贾映彬. 制造业数字化转型驱动因素研究 ——基于机器学习的证据[D]. 甘肃省兰州市. 兰州财经大学,2024.

条目包含的文件		下载所有文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
2021000009007.pdf（1766KB）	学位论文		开放获取	CC BY-NC-SA	浏览下载