Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning Approach
Effective credit risk assessment of heavy-polluting enterprises can achieve a balance between environmental and economic benefits. It requires the consideration of risk indicators for both the carbon information dimension and the compliance dimension. However, as the feature dimensions of the model...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-08-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/11/16/3462 |
| _version_ | 1827729382141067264 |
|---|---|
| author | Wanying Song Jian Min Jianbo Yang |
| author_facet | Wanying Song Jian Min Jianbo Yang |
| author_sort | Wanying Song |
| collection | DOAJ |
| description | Effective credit risk assessment of heavy-polluting enterprises can achieve a balance between environmental and economic benefits. It requires the consideration of risk indicators for both the carbon information dimension and the compliance dimension. However, as the feature dimensions of the model continue to increase, so does the irrelevant feature or noise. Therefore, we investigate the use of non-integers for regularization from high-dimensional data under the conditions of a large number of irrelevant features. In this paper, a novel Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning (WPDL) method for credit risk assessment is proposed, which could provide a sparse solution. The Wide-<i>ℓ<sub>p</sub> Penalty</i> component allows feature selection using a linear model with an <i>ℓ<sub>p</sub> Penalty</i> regularization mechanism, where 0 < <i>p</i> ≤ 2. The deep component is a DNN that can generalize indicator features from the credit risk data. The experimental results show that the minimum prediction error occurs at a non-integer <i>ℓ<sub>p</sub> Penalty</i>. Furthermore, the WPDL outperforms other models such as KNN, DT, RF, SVM, MLP, DNN, Gradient Boosting, and Bagging. |
| first_indexed | 2024-03-10T23:47:04Z |
| format | Article |
| id | doaj.art-fca5a30d21cd4e39b6fcabc308c78cd4 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-10T23:47:04Z |
| publishDate | 2023-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-fca5a30d21cd4e39b6fcabc308c78cd42023-11-19T02:02:19ZengMDPI AGMathematics2227-73902023-08-011116346210.3390/math11163462Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning ApproachWanying Song0Jian Min1Jianbo Yang2School of Management, Wuhan University of Technology, Wuhan 430070, ChinaSchool of Management, Wuhan University of Technology, Wuhan 430070, ChinaAlliance Manchester Business School, The University of Manchester, Manchester M15 6PB, UKEffective credit risk assessment of heavy-polluting enterprises can achieve a balance between environmental and economic benefits. It requires the consideration of risk indicators for both the carbon information dimension and the compliance dimension. However, as the feature dimensions of the model continue to increase, so does the irrelevant feature or noise. Therefore, we investigate the use of non-integers for regularization from high-dimensional data under the conditions of a large number of irrelevant features. In this paper, a novel Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning (WPDL) method for credit risk assessment is proposed, which could provide a sparse solution. The Wide-<i>ℓ<sub>p</sub> Penalty</i> component allows feature selection using a linear model with an <i>ℓ<sub>p</sub> Penalty</i> regularization mechanism, where 0 < <i>p</i> ≤ 2. The deep component is a DNN that can generalize indicator features from the credit risk data. The experimental results show that the minimum prediction error occurs at a non-integer <i>ℓ<sub>p</sub> Penalty</i>. Furthermore, the WPDL outperforms other models such as KNN, DT, RF, SVM, MLP, DNN, Gradient Boosting, and Bagging.https://www.mdpi.com/2227-7390/11/16/3462wide and deep learning<i>ℓ<sub>p</sub> Penalty</i>feature selectionnon-integer regularizationcredit risk assessment |
| spellingShingle | Wanying Song Jian Min Jianbo Yang Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning Approach Mathematics wide and deep learning <i>ℓ<sub>p</sub> Penalty</i> feature selection non-integer regularization credit risk assessment |
| title | Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning Approach |
| title_full | Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning Approach |
| title_fullStr | Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning Approach |
| title_full_unstemmed | Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning Approach |
| title_short | Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-<i>ℓ<sub>p</sub> Penalty</i> and Deep Learning Approach |
| title_sort | credit risk assessment of heavy polluting enterprises a wide i l sub p sub penalty i and deep learning approach |
| topic | wide and deep learning <i>ℓ<sub>p</sub> Penalty</i> feature selection non-integer regularization credit risk assessment |
| url | https://www.mdpi.com/2227-7390/11/16/3462 |
| work_keys_str_mv | AT wanyingsong creditriskassessmentofheavypollutingenterprisesawideilsubpsubpenaltyianddeeplearningapproach AT jianmin creditriskassessmentofheavypollutingenterprisesawideilsubpsubpenaltyianddeeplearningapproach AT jianboyang creditriskassessmentofheavypollutingenterprisesawideilsubpsubpenaltyianddeeplearningapproach |