Multi-task Envisioning Transformer-based Autoencoder for Corporate Credit Rating Migration Early Prediction

被引:0
|
作者
Yue, Han [1 ]
Xia, Steve [2 ]
Liu, Hongfu [1 ]
机构
[1] Brandeis Univ, Waltham, MA 02254 USA
[2] Guardian Life Insurance, New York, NY USA
关键词
Rating Migration; Fin-tech; Machine Learning; VOLATILITY; FINTECH;
D O I
10.1145/3534678.3539098
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Corporate credit ratings issued by third-party rating agencies are quantified assessments of a company's creditworthiness. Credit Ratings highly correlate to the likelihood of a company defaulting on its debt obligations. These ratings play critical roles in investment decision-making as one of the key risk factors. They are also central to the regulatory framework such as BASEL II in calculating necessary capital for financial institutions. Being able to predict rating changes will greatly benefit both investors and regulators alike. In this paper, we consider the corporate credit rating migration early prediction problem, which predicts the credit rating of an issuer will be upgraded, unchanged, or downgraded after 12 months based on its latest financial reporting information at the time. We investigate the effectiveness of different standard machine learning algorithms and conclude these models deliver inferior performance. As part of our contribution, we propose a new Multi-task Envisioning Transformer-based Autoencoder (META) model to tackle this challenging problem. META consists of Positional Encoding, Transformer-based Autoencoder, and Multi-task Prediction to learn effective representations for both migration prediction and rating prediction. This enables META to better explore the historical data in the training stage for one-year later prediction. Experimental results show that META outperforms all baseline models.
引用
收藏
页码:4452 / 4460
页数:9
相关论文
共 50 条
  • [41] VRoC: Variational Autoencoder-aided Multi-task Rumor Classifier Based on Text
    Cheng, Mingxi
    Nazarian, Shahin
    Bogdan, Paul
    WEB CONFERENCE 2020: PROCEEDINGS OF THE WORLD WIDE WEB CONFERENCE (WWW 2020), 2020, : 2892 - 2898
  • [42] A MIMO Channel Prediction Scheme Based on Multi-Task Learning
    Li, Jing
    Sun, DeChun
    Liu, ZuJun
    WIRELESS PERSONAL COMMUNICATIONS, 2020, 115 (03) : 1869 - 1880
  • [43] A Multi-Task and Transfer Learning based Approach for MOS Prediction
    Tian, Xiaohai
    Fu, Kaiqi
    Gao, Shaojun
    Gu, Yiwei
    Wang, Kai
    Li, Wei
    Ma, Zejun
    INTERSPEECH 2022, 2022, : 5438 - 5442
  • [44] A MIMO Channel Prediction Scheme Based on Multi-Task Learning
    Jing Li
    DeChun Sun
    ZuJun Liu
    Wireless Personal Communications, 2020, 115 : 1869 - 1880
  • [45] Enhancing Explainable Recommendations: Integrating Reason Generation and Rating Prediction through Multi-Task Learning
    Zhu, Xingyu
    Xia, Xiaona
    Wu, Yuheng
    Zhao, Wenxu
    APPLIED SCIENCES-BASEL, 2024, 14 (18):
  • [46] An Approach to Corporate Credit Rating Prediction Using Computational Intelligence-Based Methods
    Zukanovic, Milica
    Milosevic, Pavle
    Poledica, Ana
    Vucicevic, Aleksandra
    SUSTAINABLE BUSINESS MANAGEMENT AND DIGITAL TRANSFORMATION: CHALLENGES AND OPPORTUNITIES IN THE POST-COVID ERA, 2023, 562 : 80 - 95
  • [47] Credit scoring using multi-task Siamese neural network for improving prediction performance and stability
    Kwon, Soonjae
    Jang, Jaeyeon
    Kim, Chang Ouk
    EXPERT SYSTEMS WITH APPLICATIONS, 2025, 259
  • [48] Autism spectrum disorders detection based on multi-task transformer neural network
    Gao, Le
    Wang, Zhimin
    Long, Yun
    Zhang, Xin
    Su, Hexing
    Yu, Yong
    Hong, Jin
    BMC NEUROSCIENCE, 2024, 25 (01):
  • [49] Multi-Task Mean Teacher Medical Image Segmentation Based on Swin Transformer
    Zhang, Jie
    Li, Fan
    Zhang, Xin
    Cheng, Yue
    Hei, Xinhong
    APPLIED SCIENCES-BASEL, 2024, 14 (07):
  • [50] Multi-Modal Pedestrian Crossing Intention Prediction with Transformer-Based Model
    Wang, Ting-Wei
    Lai, Shang-Hong
    APSIPA TRANSACTIONS ON SIGNAL AND INFORMATION PROCESSING, 2024, 13 (05)