Ph.D Candidate
South China University of Technology (Research Affiliate)
Research assistent
Guangzhou Financial Service Innovation and Risk Management Research Base
Research and Development Engineer
Dongguan Yong Hua Hao Information Technology Ltd.
Contact:
Department of School of Business Administration
South China University of Technology
381 Wushan Road, Guangzhou
GuangDong, China
Hosted on GitHub Pages — Theme by orderedlist
I am currently an Ph.d candidate of Management Science and Engineering at the South China University of Technology and Research assistent at the Guangzhou Financial Service Innovation and Risk Management Research Base. I graduated with a Master in Information Technology from The University of Melbourne in July 2018
My research interests include asset pricing, complex financial systems, systemic financial risk, and green finance.
Unveiling Systemically Important Corporations: A Dynamic Multilayer Financial Network Approach
with Yue Zhang, Xiaolei He
R & R, International Review of Economics and Finance (updated July 28, 2024)
This study aims to evaluate the systemic importance of corporations in the Chinese A-share market using dynamic multilayer complex networks and multi-criteria decision-making (MCDM) techniques. The dynamic multilayer financial network consists of a tail risk network established with quantile regression and a trading information network constructed with principal component analysis. Regarding network structure, the topological indicators of multilayer networks are more sensitive to systemic financial risk than those of single-layer networks. Subsequently, by combining multidimensional complex network theory with the Technique for Order Preference by Similarity to the Ideal Solution (TOPSIS), the systemic importance evaluation (SIE) is evaluated.
A Study on Systematic Risk Measurement in Chinese Foreign Exchange Market Based on Knowledge Graph Network Features
with Weiguo Zhang, Jiaxian Shen
R & R, Journal of Management Science(《管理科学学报》) (updated July 28, 2024)
Systemic risk in the foreign exchange market not only affects the healthy development of the foreign exchange market, but also influences the stability of the whole financial system. Considering the systemic importance role of foreign exchange market makers in the foreign exchange market, we analyse the mechanism of systemic risk formation and aggregation in China's foreign exchange market, explore the channels of systemic risk formation in the foreign exchange market under negative shocks, and construct a systemic risk measurement index for China's foreign exchange market by combining the network features of the Knowledge Graph with the existing three types of systemic risk measurement indexes, ΔCoVaR, MES, and SRISK. The validity of the indicators is verified through major events at home and abroad, the systemic importance of the foreign exchange market is identified as foreign exchange market maker banks, and the impact characteristics under different shock channels are explored. The results of the study show that: 1) MES can reflect the changes in the level of systemic risk in the foreign exchange market under short-event shocks, i.e., its short-term measurement is the best, while SRISK can reflect the overall changes in the level of systemic risk in the foreign exchange market under long-event shocks, i.e., its long-term measurement is the most effective; 2) the indicator measurement under the consideration of the network features of the Knowledge Graph is better than that without the use of the network features, and the use of the CFETS index as the index of China's foreign exchange market is better than the SDR index and BIS index; 3) In the long run, the systemic importance of the four major state-owned banks and the Bank of Communications is higher, and in the short run, the systemic importance of the local banks is higher; 4) In the case of a long-event shock, it is necessary to strengthen the regulation of the banks with a high systemic importance, and the regulatory level of banks with a low systemic importance can be appropriately adjusted downward when considering the regulatory cost. supervision level for banks of lower importance.
A Knowledge Graph-based Approach to Identifying Financial Fraud in Listed Companies
with Weiguo Zhang, Yuemin Yang, et al
R & R, Management Review(《管理评论》) (updated July 28, 2024)
Financial fraud of listed companies poses a significant threat to the stability of the financial market, and related studies that have existing financial fraud identification models for listed companies seldom consider the associative features between listed companies and their relationships with financial entities, and it is to be explored whether these relationships can improve the accuracy of fraud identification. This study proposes a financial fraud identification method that considers the associative relationships of listed companies, which uses knowledge mapping to portray common personnel, concerted action, and common related party relationships among listed companies, extracts isomorphic networks through knowledge reasoning, combines isomorphic network features with the company's own features, and constructs a fraud identification model using the Lightweight Gradient Boosting Machine algorithm (LightGBM). Based on the real data of China's A-share listed companies from 2016 to 2020 for empirical testing, the results show that the associative relationship network features help to improve the accuracy of financial fraud identification, and LightGBM combining network features achieves the best performance compared with other machine learning methods.
A study on dynamic financial distress prediction based on RandomForest model: from financial network perspective
with Yue Zhang, Yuan Zhao
R & R, JKnowledge Based System (updated July 28, 2024)
As the extent of interdependence between sectors and institutions within the financial system escalates, the insolvency of enterprises typically results in adverse consequences for the financial standing of affiliated companies. Further corporate bankruptcies can occur in this process's market-wide cascading effect. In light of this, the intricate interconnectedness characteristics among businesses are a current research hot topic regarding predicting corporate financial distress. This research aims to construct a novel hybrid model incorporating a network-characterized multidimensional financial distress indicator system to forecast corporate financial distress dynamically. The first steps of our work begin with constructing a volatility spillover network and extracting network-based features utilizing the DCC-GARCH model. Subsequently, experiments are executed utilizing the hybrid model ADAPOS-RF, which is constructed utilizing Random Forest, AdaBoost, and PSO for parameter optimization. The empirical results indicate that when applied to a hybrid feature dataset, the proposed ADAPSO-RF model exhibits superior predictive power compared to the Random Forest series of benchmark models. Moreover, compared to widely employed machine learning models, the accuracy performance of the proposed model surpasses that of the worst model by 36.6\%. In an economic sense, our research can offer timely indicators of corporate financial downturns to business operators, investors, and government regulators, facilitating the development and enhancement of risk prevention strategies.
Research on the effect of firm-specific investor sentiment on the idiosyncratic volatility anomaly: Evidence from the Chinese market
with Yue Zhang.
Pacific-Basin Finance Journa, 2023,102114.
This paper evaluates how investor sentiment contributes to explaining the idiosyncratic volatility (IVOL) anomaly from a firm-level perspective in the Chinese stock market. This study employs the principal component analysis method to construct firm-specific investor sentiment (FSIS) and analyzes the effects of investor sentiment on the IVOL anomaly. In contrast to market-level sentiment, FSIS, as a varying cross-sectional sentiment indicator, can contribute to interpreting the idiosyncratic volatility anomaly from a micro- and portfolio-level viewpoint. According to our empirical investigation, stocks with higher FSIS have lower IVOL returns, i.e., a significant positive correlation exists between FSIS and IVOL.
@article{chen2023research,
title={Research on the effect of firm-specific investor sentiment on the idiosyncratic volatility anomaly: Evidence from the Chinese market},
author={Chen, Haozhi and Zhang, Yue},
journal={Pacific-Basin Finance Journal},
pages={102114},
year={2023},
publisher={Elsevier}
}
The Research and Application of System for Paper Surface Morphology Stereoscopic Observation and Characterization
with Tingli Liu, Guangxue Chen.
International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems, 2021, 29(Supp01): 89–100.
The work aims to explore a microscopic observation system of paper surface and achieve high-precision stereoscopic observation with detail characterization of paper surface morphology. Based on the DT-400E precision program-controlled three-dimensional translation stage and KEYENCE LJV-7200 two-dimensional laser scanner, the hardware parts of our own system are developed to scan and transmit point cloud data of paper surface morphology to the computer. The corresponding system software will automatically process the point cloud data acquired from the laser scanner and generate the corresponding vivid 3D model and height histogram. This system scans and characterizes four different types of paper samples, allowing the human eye to visually distinguish the differences in surface morphology as well as automatically calculate the numerical differences in paper surface morphology parameters. The results of the applicability test show that the system is highly efficient in acquiring, observing, and evaluating the topography of the paper surface. The system can not only predict the paper surface quality of printed paper, but can also be extended to the evaluation of 3D printed surfaces.
@article{chen2021research,
title={The Research and Application of System for Paper Surface Morphology Stereoscopic Observation and Characterization},
author={Chen, Haozhi and Liu, Tingli and Chen, Guangxue},
journal={International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems},
volume={29},
number={Supp01},
pages={89--100},
year={2021},
publisher={World Scientific}
}
Naked-eye 3D display based on microlens array using combined micro-nano imprint and UV offset printing methods
with Linyi Chen, Guangxue Chen, Liyu Liao.
Molecules, 2020, 25(9): 2012.
An optical film integrating microlens array (MLAs) and 3D micro-graphics is an important way to achieve the naked-eye 3D display effect. The 3D micro-graphics is traditionally generated by the micro-nano imprint technology based on precision engraving mold, which leads to high production cost and low production efficiency, and thus restricts the rapid response to production tasks and large-scale popularization and application. In this study, a process scheme for large-scale printing of 3D micro-graphics using UV offset printing based on presensitized (PS) plate was proposed, matching with the MLAs fabricated by micro-nano imprint process to achieve naked-eye 3D display effect. We used the laser confocal microscope to systematically measure and analyze the geometric and optical performance of the fabricated MLAs in terms of height, curvature radius, center distance, spacing, focal length, and numerical aperture, and evaluated the influence of the publishing resolution of the PS plate on the display effect of 3D micro-graphics. The printing quality and display effect of 3D micro-graphics were further improved by adjusting process parameters such as printing speed and printing pressure. The results of the current study demonstrate that the combined application of micro-nano imprint technology based on precision mold and UV offset printing technology based on PS plate can achieve an excellent naked-eye 3D display effect in 360° all angles, which is efficient, cost-saving, and highly flexible.
@article{chen2020naked,
title={Naked-eye 3D display based on microlens array using combined micro-nano imprint and UV offset printing methods},
author={Chen, Linyi and Chen, Guangxue and Liao, Liyu and Chen, Haozhi},
journal={Molecules},
volume={25},
number={9},
pages={2012},
year={2020},
publisher={MDPI}
}
Realization of Rapid Large-Size 3D Printing Based on Full-Color Powder-Based 3DP Technique
with Guangxue Chen, Xiaochun Wang, Chen Chen.
Molecules, 2020, 25(9): 2037.
The powder-based 3DP (3D printing) technique has developed rapidly in creative and customized industries on account of it’s uniqueness, such as low energy consumption, cheap consumables, and non-existent exhaust emissions. Moreover, it could actualize full-color 3D printing. However, the printing time and size are both in need of upgrade using ready printers, especially for large-size 3D printing objects. Given the above issues, the effects of height and monolayer area on printing time were explored and the quantitative relationship was given in this paper conducted on the specimens with a certain gradient. On this basis, an XYX rotation method was proposed to minimize the printing time. The mechanical tests were conducted with three impregnation types as well as seven printing angles and combined with the characterization of surface structure based on the scanning electron microscope (SEM) digital images to explore the optimum parameters of cutting-bonding frame (CBF) applied to powder-based 3D printing. Then, four adhesives were compared in terms of the width of bonded gap and chromatic aberration. The results revealed that ColorBond impregnated specimens showed excellent mechanical properties which reached maximum when printed at 45° to Z axis, and α-cyanoacrylate is the most suitable adhesive to bond full-color powder-based models. Finally, an operation technological process was summarized to realize the rapid manufacturing of large-size full-color 3D printed objects.
@article{chen2020realization,
title={Realization of Rapid Large-Size 3D Printing Based on Full-Color Powder-Based 3DP Technique},
author={Chen, Guangxue and Wang, Xiaochun and Chen, Haozhi and Chen, Chen},
journal={Molecules},
volume={25},
number={9},
pages={2037},
year={2020},
publisher={MDPI}
}
3D Reproduction of Oil Painting Based on UV Ink-Jet Technology
with Yansong Peng, Chen Chen, Liuxi He.
2015 International Conference on Electromechanical Control Technology and Transportation, Atlantis Press, 2015: 368-372..
The goal of this experiment is to use UV ink as the modeling material to reproduce oil paintings with colorful three-dimensional features. The scanned picture of oil painting stored in bitmap format was sent into a software program that provided stratification processing. The layers were printed on paper or canvas using 3D print technology in the UV-LED ink-jet printer, with a sequence of positive and reverse outputs, and two output methods consisting of a combination of multilayer white ink with a layer of colored ink, and a combination of a layer of colored ink with multilayer transparent ink. The results indicated that replicating oil paintings with fine three-dimensional details and vivid color effects could be realized when the layers after stratification were printed out using a reverse sequence of multilayer white ink and colored ink. Therefore, UV ink-jet technology can successfully realize the reproduction of oil paintings with colorful three-dimensional features.
@inproceedings{peng20153d,
title={3D Reproduction of Oil Painting Based on UV Ink-Jet Technology},
author={Peng, Yansong and Chen, Chen and He, Liuxi and Chen, Haozhi},
booktitle={2015 International Conference on Electromechanical Control Technology and Transportation},
pages={368--372},
year={2015},
organization={Atlantis Press}
}
Evaluation of Technology Projects using Multi-level Criteria and TOPSIS Approach
with Bing Mo, Bin Shang, Baimao Lei, Guangxue Chen.
WCSE 2022 Spring Event: 2022 9th International Conference on Industrial Engineering and Applications.
The evaluation of technology projects for commercialization is intrinsically difficult but of great importance. In this paper, a project evaluation criteria and model are constructed in the following two aspects. Firstly, multi-level and multi-dimensional project evaluation criteria are formed with reference to various studies. These criteria help decision-makers to gain a comprehensive understanding of the technical level, maturity, implementability, promotion value, and industrialization foundation of the project. Secondly, since the evaluation of the project involves multiple criteria, it can be modeled as a multiple criteria decision making (MCDM) problem. A nested MCDM model for multi-level criteria is constructed using the technique for order preference by similarity to ideal solution (TOPSIS) algorithm with a hybrid of AHP and Entropy weighting methods. The results illustrate the effectiveness of the model in evaluating technology projects. The outcomes of the research enable decision-makers to comprehensively understand the projects based on multi-level criteria while making practical and rational decisions based on the evaluation model.
@article{moevaluation,
title={Evaluation of Technology Projects using Multi-level Criteria and TOPSIS Approach},
author={Mo, Bing and Chen, Haozhi and Shang, Bin and Lei, Baimao and Chen, Guangxue}
}
Early warning software for corporate financial distress based on financial knowledge graph network structure
with Weiguo Zhang, Xiaowei Yang.
The purpose of this software development is to provide financial regulators and investors with an early warning function of financial distress of listed companies (enterprises) and to avoid systemic risk.
2024SR0360473.
Software for corporate bond default risk assessment based on financial knowledge graph network structure
with Weiguo Zhang, Xiaowei Yang.
The purpose of this software development is to provide financial regulators and investors with a systemic importance ranking of listed companies and to clarify their risk levels.
2024SR0792724.
Systematic Importance Assessment Software for Listed Companies Based on Financial Knowledge Graph Network Structure
with Weiguo Zhang.
The purpose of this software development is to provide financial regulators and investors with an assessment of the default risk in the corporate bond market and to minimize the probability of systemic risk.
2024SR0177829.
A kind of temporary signal light device
with Guangxue Chen, Linyi Chen.
The temporary signal light device, through the bottom of the bottom frame upper and lower drive cylinder settings, can drive the bottom of the load-bearing block and the universal wheel at the same time close to the ground, to ensure that there are five contact points, while ensuring that it is always not isotropic, to increase the stability of the placement, and at the same time without manual adjustment, the degree of automation is higher, easy to place, and at the same time to ensure that the solar energy utilization, solving the existing device stability and stability of the device, the bottom side of the top frame is fixedly mounted on the top side of the top frame. Utilization rate, solving the problem of limited stability of the existing device and lower efficiency of solar energy utilization.
CN218918259U.
An underground parking lot waterproofing device
with Guangxue Chen, Linyi Chen.
The utility model discloses an underground parking lot waterproofing device, relates to the field of auxiliary waterproofing for parking lots, including a door seat, said door seat is provided with a drain groove in a rectangular shape, said door seat is provided with a retaining slot on the top side of said door seat, said retaining groove is provided with a baffle plate set in said retaining groove, and a conduction groove which is connected to the internal cavity of the drain groove is provided on the bottom surface of the inside of the retaining groove; this underground parking lot waterproofing device, by the installation of various components on the inside of the door seat, when a small part of the rainwater to be immersed in the door seat, the provision of the drain slot on the baffle plate on the top side of the door sill is able to convey a small part of the rainwater to the drain slot for temporary storage, and the underground drainage pipe conducts it, and when the majority of the rainwater is to be immersed in the door sill, the baffle plate is rotated so that the baffle slot is exposed to the surface of the sill, and at this time, most of the rainwater automatically flows to the baffle slot, and passes through the conduction slot to enter the drain slot, and finally, sewage is discharged by the underground drain on the bottom side of the sill. Finally, the sewage will be discharged by the underground drainage pipe on the bottom side, which is simple in structure and has a certain amount of water storage, and is suitable for efficiently dredging and discharging different amounts of rainwater according to the demand.
CN218991429U.
A kind of road marking equipment
with Guangxue Chen, Linyi Chen.
The utility model discloses a road marking equipment, relates to the field of road marking, including a base plate, the top of said base plate is fixedly mounted with a paint bucket, the output end of said paint bucket is fixedly mounted with a pump body, and the output end of the pump body is fixedly connected with a corrugated pipe; the top of said front set of seats is fixedly mounted with a support that extends backward, the top of said support is fixedly mounted with a heater fan, and the bottom side of the support is equipped with a wind shield for the output end of the heater fan. The road marking equipment, through the setup of each component on the base plate, when the electric telescopic rod drives the rear sleeve plate and the front sleeve plate to extend outward to adjust the position of the rear sleeve seat, then the front sleeve seat is also adjusted accordingly, which can ensure that the wind cover is always on the bottom side of the spraying seat to spray the real-time drying of the marking line, and the setup of the inner side of the paint bucket filled with the corresponding curing agent, which is able to spray the line after the curing effect, solving the problem of the existing device, which is the best solution to the problem of the existing device, which is the best solution to the problem of the existing device. Curing effect, to solve the existing device spraying marking line after the early drying of slow and easy to be contaminated.
CN219059717U.
A carbon fiber paper with directional water transport function and its preparation method and application
with Guangxue Chen, Zihan Jiang, Junfei Tian, Xiaofang Wan, Qifeng Chen, Minghui He.
The present invention belongs to the technical field of fuel cell materials, and discloses a carbon fiber paper with a directional water transport function and a preparation method and application thereof. Said preparation method includes the following steps: dispersing polytetrafluoroethylene and deionized water uniformly to obtain a polytetrafluoroethylene emulsion; treating carbon paper using the polytetrafluoroethylene emulsion to obtain a hydrophobic carbon paper; and subjecting one side of the obtained hydrophobic carbon paper to plasma treatment to produce carbon fiber paper with a directional water transport function. The preparation method of the present invention is simple, easy to operate, and suitable for industrialized production; the product can enhance the gas drainage function of the gas diffusion layer of the proton exchange membrane fuel cell, effectively prevent the occurrence of water flooding phenomenon, and enhance the strength of the carbon fiber paper, which is conducive to the subsequent processing of the gas diffusion layer, and it has a good application prospect.
CN115075056A.
A method of preparing a gas diffusion layer of a proton exchange membrane fuel cell and its products
with Guangxue Chen, Zihan Jiang, Junfei Tian, Xiaofang Wan, Qifeng Chen, Qi Zhu.
The present invention relates to a method for preparing a gas diffusion layer for a proton exchange membrane fuel cell and a product thereof, comprising coating a slurry of polydopamine grafted micronized carbon particles on the surface of a carbon paper by a rapid printing or coating technique; preparing a thin layer of graphene oxide on the surface of the carbon paper by layer-by-layer self-assembly using graphene oxide flakes; assembling to form a continuous mono- or multilayered thin film on the surface of the carbon paper using a nanocarbon spheres slurry; using a PTFE emulsion to treat the gas diffusion layer to form a low surface energy thin layer for deposition. Using polydopamine-modified conductive carbon black, graphene oxide and carbon nanosphere particles, there are differences in geometric characteristics between the three materials, layer by layer coating to form a gradient change in pore size structure, and at the same time, PTFE treatment to form a hydrophobic structure, through the construction of a gradient GDL to improve the gas diffusion layer conductivity of water, gas and improve the efficiency of the electrochemical reaction. The product carbon paper maintains good mechanical strength and is easy for subsequent processing and surface treatment operations.
CN114864966A.