Interpretable Machine Learning Framework for Corporate Financialization Prediction: A SHAP-Based Analysis of High-Dimensional Data
Abstract
1. Introduction
- We develop a comprehensive interpretable machine learning framework that systematically integrates SHAP analysis with high-dimensional feature selection and multi-algorithm optimization, specifically tailored for corporate financialization prediction.
- We develop an advanced methodological pipeline that integrates nine optimized machine learning algorithms with Hippopotamus Optimization (HO) for systematic hyperparameter tuning. The framework incorporates SHAP analysis for mathematically grounded feature importance quantification and interaction effect visualization.
- We establish a comprehensive evaluation framework for assessing explainable machine learning performance in financial applications. The methodology demonstrates robustness across distinct economic phases and provides systematic approaches for cross-validation, sensitivity analysis, and interpretability validation.
2. Preliminaries
2.1. Data and Sample
2.2. Variable Definitions
3. Methodology
3.1. Framework and Model Structure
Algorithm 1: Corporate Financialization Prediction |
Initialization: |
1. Load dataset of non-financial Chinese A-share listed companies from 2011 to 2022 |
2. Data preprocessing: |
- Exclude financial firms, ST and *ST firms |
- Handle missing data: fill continuous variables with their mean, complete discrete variables with their plural |
- Winsorize continuous variables (top 1% and bottom 1%) |
3. Split the data into training (80%) and test (20%) sets |
4. Define independent variables: |
- 19 CSR-related features, including the overall CSR score, primary, secondary, and tertiary CSR dimensions |
- 21 financialization-related variables |
5. Lag features by 1 year to account for reverse causation |
6. Perform feature selection using Lasso regression to reduce multicollinearity |
Phase 1: Model Training and Hyperparameter Optimization |
1. Train 9 machine learning models on the training dataset: |
- XGBoost, RF, LightGBM, CatBoost, AdaBoost, SVR, DT, RR, KNN |
2. Use five-fold cross-validation with HO to optimize hyperparameters: |
- For each fold, train the model with selected hyperparameters and evaluate performance |
- The hyperparameter optimization process adjusts parameters based on exploration and exploitation phases of HO algorithm |
- Calculate model performance using evaluation metrics (R2, MAE, MSE, RMSE) |
Phase 2: Model Evaluation |
1. Evaluate model performance on the test dataset: |
- Use the following performance metrics: |
- R2 = 1 − |
- |
- |
- |
2. Identify the best-performing model based on evaluation metrics (XGBoost, for example) |
Phase 3: SHAP Interpretation |
1. Use SHAP to explain model predictions: |
- For each model, calculate the Shapley values (φ) for each feature: |
- |
- Interpret SHAP feature importance to identify key drivers of corporate financialization |
2. Visualize the feature importance using SHAP summary plots and feature dependence plots |
3. Analyze the interaction effects using SHAP interaction plots to understand conditional influences of variables like EPU and CSR dimensions |
Termination: |
1. After model evaluation and SHAP interpretation, finalize the best model and report the results |
2. Provide actionable insights based on model findings for financial decision-making and policy formulation |
End |
3.2. Machine Learning Algorithms
3.3. HO
Algorithm 2: Hippopotamus Optimization (HO) |
Initialization: |
1. Define the search space bounds for each dimension: |
2. Initialize the population of hippos randomly: |
For i = 1 to N (Population size): |
For j = 1 to m (Dimensions): |
) // where r ∈ [0, 1] is a random variable |
Initialize population matrix |
Phase 1: Exploration (Aquatic Behavior) |
1. Divide hippos into two groups: |
- Dominant male hippos (Directly move towards best-known solution) |
- Female and immature hippos (Explore with diversity) |
2. Update position of male hippos (Dominant hippos): |
For i = 1 to N: |
For j = 1 to m: |
// (Random integer coefficient) |
3. Update position of female and immature hippos (Exploration with temperature-based mechanism): |
For i = 1 to N: |
If T > 0.6: |
// Exploit good regions or follow others: |
// is a perturbation factor, is the mean position of nearby hippos |
Else: |
// Random or reverse direction based on probability: |
ϕ = , if > 0.5 |
4. Greedy Selection (Solution Acceptance): |
- For males: |
If : |
Else: |
- For females/immature hippos: |
If : |
Else: |
// F(⋅) is the objective function to minimize |
Phase 2: Exploitation (Defensive Behavior) |
1. Introduce a virtual predator randomly into the search space: |
2. Evaluate distance between predator and each hippo: |
= || |
3. If predator is close: |
- Hippos move towards the predator, potentially exploiting the region around the predator: |
= + |
// Adaptive control terms are problem-specific and adjust the defense behavior |
Termination: |
1. Repeat exploration and exploitation phases until: |
- Maximum number of iterations reached, or |
- Convergence tolerance is satisfied |
2. The best solution found, , is reported as the final output. |
End |
3.4. SHAP
3.5. Performance Evaluation Criteria
4. Experimental Results
4.1. Descriptive Statistics
4.2. Performance Comparison of Machine Learning Models
- Non-linear relationships: Traditional econometric models assume linear relationships between variables, while our subsequent SHAP analysis (Section 4.6) reveals significant non-linearities, particularly U-shaped relationships for CSR dimensions.
- High-dimensional interactions: With 40 variables, the number of potential interactions is substantial. While econometric models require manual specification of interaction terms, tree-based methods like XGBoost automatically capture these complex interactions.
- Multicollinearity: The econometric models suffer from severe multicollinearity issues. Our diagnostic tests show that 18 out of 40 variables have Variance Inflation Factors (VIFs) exceeding 10, with some reaching as high as 25. In contrast, tree-based methods are inherently robust to multicollinearity due to their splitting mechanisms.
4.3. XGBoost Training Dynamics and Convergence Analysis
4.4. Variable Importance in Corporate Financialization: SHAP Analysis
4.5. SHAP Summary Plots and Their Implications for Corporate Financialization
4.6. Single-Factor Analysis Through SHAP Dependence Plots
4.7. Interaction Effects Between Key Factors and Contingent Variables: SHAP Analysis
5. Discussion
6. Conclusions
Author Contributions
Funding
Data Availability Statement
Conflicts of Interest
References
- Klinge, T.J.; Fernandez, R.; Aalbers, M.B. Whither corporate financialization? A literature review. Geogr. Compass 2021, 15, e12588. [Google Scholar] [CrossRef]
- Su, K.; Lu, Y. The impact of corporate social responsibility on corporate financialization. Eur. J. Financ. 2023, 29, 2047–2073. [Google Scholar] [CrossRef]
- Wu, K.; Lu, Y. Corporate digital transformation and financialization: Evidence from Chinese listed firms. Financ. Res. Lett. 2023, 57, 104229. [Google Scholar] [CrossRef]
- Zhang, Z.; Su, Z.; Tong, F. Does digital transformation restrain corporate financialization? Evidence from China. Financ. Res. Lett. 2023, 56, 104152. [Google Scholar] [CrossRef]
- Si, D.K.; Zhuang, J.; Ge, X.; Yu, Y. The nexus between trade policy uncertainty and corporate financialization: Evidence from China. China Econ. Rev. 2024, 84, 102113. [Google Scholar] [CrossRef]
- Sui, B.; Yao, L. The impact of digital transformation on corporate financialization: The mediating effect of green technology innovation. Innov. Green Dev. 2023, 2, 100032. [Google Scholar] [CrossRef]
- Cheng, Z.; Masron, T.A. Does economic policy uncertainty exacerbate corporate financialization? Evidence from China. Appl. Econ. Lett. 2024, 31, 1028–1036. [Google Scholar] [CrossRef]
- Du, P.; Zheng, Y.; Wang, S. The minimum wage and the financialization of firms: Evidence from China. China Econ. Rev. 2022, 76, 101870. [Google Scholar] [CrossRef]
- Jiang, F.; Shen, Y.; Cai, X. Can multiple blockholders restrain corporate financialization? Pac.-Basin Financ. J. 2022, 75, 101827. [Google Scholar] [CrossRef]
- Ma, X.; Xu, Q. The impact of carbon emissions trading policy on carbon emission of China’s power industry: Mechanism and spatial spillover effect. Environ. Sci. Pollut. Res. 2023, 30, 74207–74222. [Google Scholar] [CrossRef]
- Wang, H.; Sun, K.; Xu, S. Does housing boom boost corporate financialization?—Evidence from China. Emerg. Mark. Financ. Trade 2023, 59, 1655–1667. [Google Scholar] [CrossRef]
- Cao, W.; Chen, C.; Jiang, D.; Li, W.; Zhang, Y. Industrial policy and non-financial corporations’ financialization: Evidence from China. Eur. J. Financ. 2022, 28, 397–415. [Google Scholar] [CrossRef]
- Zhang, X.; Zheng, X. Does carbon emission trading policy induce financialization of non-financial firms? Evidence from China. Energy Econ. 2024, 131, 107316. [Google Scholar] [CrossRef]
- Liu, H.; Pan, H. Reducing carbon emissions at the expense of firm physical capital investments and growing financialization? Impacts of carbon trading policy from a regression discontinuity design. J. Environ. Manag. 2024, 356, 120577. [Google Scholar] [CrossRef] [PubMed]
- Su, K.; Zhao, Y.; Wang, Y. Customer concentration and corporate financialization: Evidence from non-financial firms in China. Res. Int. Bus. Financ. 2024, 68, 102159. [Google Scholar] [CrossRef]
- Zhong, H.; Al-Duais, Z.A.M.; Peng, B. The impact of idiosyncratic risk on corporate financialization—Evidence from China. Int. Rev. Financ. Anal. 2023, 86, 102491. [Google Scholar] [CrossRef]
- Hou, Q.; Tang, X.; Teng, M. Labor costs and financialization of real sectors in emerging markets. Pac.-Basin Financ. J. 2021, 67, 101547. [Google Scholar] [CrossRef]
- Tang, Y.; Wang, L.; Shu, H. Tax reduction and the financialization of real enterprises: Evidence from China’s VAT reform. Int. Rev. Econ. Financ. 2024, 92, 835–850. [Google Scholar] [CrossRef]
- Zhang, Y.; Zhang, H.; Yang, L.; Xu, P. Managerial ownership and corporate financialization. Financ. Res. Lett. 2023, 58, 104682. [Google Scholar] [CrossRef]
- Gao, R.; Cui, S.; Wang, Y.; Xu, W. Predicting financial distress in high-dimensional imbalanced datasets: A multi-heterogeneous self-paced ensemble learning framework. Financ. Innov. 2025, 11, 50. [Google Scholar] [CrossRef]
- Elena, D.; Hué, S.; Hurlin, C.; Tokpavi, S. Machine learning for credit scoring: Improving logistic regression with non-linear decision-tree effects. Eur. J. Oper. Res. 2022, 297, 1178–1192. [Google Scholar]
- Liu, J.; Liu, Z.; Li, Q.; Kong, W.; Li, X. Multi-Domain Controversial Text Detection Based on a Machine Learning and Deep Learning Stacked Ensemble. Mathematics 2025, 13, 1529. [Google Scholar] [CrossRef]
- Li, X.; Wang, X.; Kong, T.; Zheng, J.; Luo, M. From bitcoin to solana–innovating blockchain towards enterprise applications. In Proceedings of the International Conference on Blockchain, Virtual Event, 10–14 December 2021; Springer International Publishing: Berlin/Heidelberg, Germany; pp. 74–100. [Google Scholar]
- Li, X.; Sigov, A.; Ratkin, L.; Ivanov, L.A.; Li, L. Artificial intelligence applications in finance: A survey. J. Manag. Anal. 2023, 10, 676–692. [Google Scholar] [CrossRef]
- Campisi, G.; Muzzioli, S.; De Baets, B. A comparison of machine learning methods for predicting the direction of the US stock market on the basis of volatility indices. Int. J. Forecast. 2024, 40, 869–880. [Google Scholar] [CrossRef]
- Jones, S.; Moser, W.J.; Wieland, M.M. Machine learning and the prediction of changes in profitability. Contemp. Account. Res. 2023, 40, 2643–2672. [Google Scholar] [CrossRef]
- Li, Z. Extracting spatial effects from machine learning model using local interpretation method: An example of SHAP and XGBoost. Comput. Environ. Urban Syst. 2022, 96, 101845. [Google Scholar] [CrossRef]
- Martins, T.; De Almeida, A.M.; Cardoso, E.; Nunes, L. Explainable artificial intelligence (XAI): A systematic literature review on taxonomies and applications in finance. IEEE Access 2023, 12, 618–629. [Google Scholar] [CrossRef]
- Wang, S.; Meng, H.; Chen, Z. Exploring the links between bank competition, economic policy uncertainty, and corporate financialization. Financ. Res. Lett. 2025, 78, 107100. [Google Scholar] [CrossRef]
- Wang, M.; Mohd Nor, N.; Abdul Rahim, N.; Khan, F.; Zhou, Z. Trade policy uncertainty and corporate financialization: Strategic implications for non-financial firms in China. Cogent Econ. Financ. 2025, 13, 2460078. [Google Scholar] [CrossRef]
- Shen, J.; Xu, M.; Liu, X.; Zhao, Y. Active adaptation or short-run profit pursuing? Carbon emissions trading and corporate financialization: Evidence from Chinese listed companies. Environ. Dev. Sustain. 2024, 1–18. [Google Scholar] [CrossRef]
- Gao, C.; Zhang, S. ESG performance and corporate financialization: A dual perspective of risk management and value creation. Financ. Res. Lett. 2025, 71, 106442. [Google Scholar] [CrossRef]
- Xu, S. Corporate social responsibility and financialization: Is CSR used as a financial tool? Eur. Manag. Rev. 2025, 1–19. [Google Scholar] [CrossRef]
- Zhang, J.; Ma, X.; Zhang, J.; Sun, D.; Zhou, X.; Mi, C.; Wen, H. Insights into geospatial heterogeneity of landslide susceptibility based on the SHAP-XGBoost model. J. Environ. Manag. 2023, 332, 117357. [Google Scholar] [CrossRef] [PubMed]
- Sun, Y.; Qu, Z.; Liu, Z.; Li, X. Hierarchical Multi-Scale Decomposition and Deep Learning Ensemble Framework for Enhanced Carbon Emission Prediction. Mathematics 2025, 13, 1924. [Google Scholar] [CrossRef]
- Amiri, M.H.; Hashjin, N.M.; Montazeri, M.; Mirjalili, S.; Khodadadi, N. Hippopotamus optimization algorithm: A novel nature-inspired optimization algorithm. Sci. Rep. 2024, 14, 5032. [Google Scholar] [CrossRef] [PubMed]
- Zhang, Y.; Guo, J.; Ding, X.; Tang, Z.; Wang, T.; Wu, W.; Jia, W. Optimizing Communication Efficiency through Training Potential in Multi-Modal Federated Learning. ACM Trans. Internet Technol. 2025. [Google Scholar] [CrossRef]
Variable | Obs | Mean | Std | Min | Median | Max |
---|---|---|---|---|---|---|
Fin_lagged | 25,642 | 0.07 | 0.1 | 0 | 0.03 | 0.51 |
SOE | 25,642 | 0.32 | 0.47 | 0 | 0 | 1 |
SA | 25,642 | −3.78 | 0.25 | −4.4 | −3.79 | −3.11 |
Carbontrade | 25,642 | 0.13 | 0.33 | 0 | 0 | 1 |
Digital | 25,642 | 1.46 | 1.42 | 0 | 1.1 | 5.21 |
EPU | 25,642 | 4.25 | 2.29 | 1.14 | 3.65 | 7.92 |
Lottery | 25,642 | 0.44 | 0.14 | 0.2 | 0.44 | 0.9 |
PLD_DUM | 25,642 | 0.41 | 0.49 | 0 | 0 | 1 |
IP | 25,642 | 0.09 | 0.29 | 0 | 0 | 1 |
Multi | 25,642 | 0.26 | 0.44 | 0 | 0 | 1 |
MA | 25,642 | 0.54 | 0.28 | 0 | 0.65 | 0.99 |
Customer | 25,642 | 0.31 | 0.22 | 0 | 0.26 | 1.58 |
Minwage | 25,642 | 1668.49 | 306.99 | 880 | 1668.49 | 2360 |
TPU | 25,642 | 2.3 | 0.41 | 1.65 | 2.44 | 2.84 |
Lnhp | 25,642 | 13,126.13 | 9670.75 | 3553.96 | 10,542.48 | 46,941 |
Media | 25,642 | 4.36 | 1.08 | 0.69 | 4.39 | 7.27 |
Top1 | 25,642 | 35.08 | 14.63 | 9.31 | 33.08 | 74.82 |
Mshare | 25,642 | 16.81 | 20.99 | 0 | 4.33 | 89.99 |
DIF | 25,642 | 2.85 | 1.05 | 0.33 | 2.98 | 4.59 |
HHI | 25,642 | 0.91 | 0.09 | 0.41 | 0.94 | 0.99 |
Marketization | 25,642 | 7.55 | 4.37 | 0 | 9.58 | 12.39 |
Firmsize | 25,642 | 22.17 | 1.32 | 19.29 | 21.95 | 28.64 |
CSR Total rating | 25,642 | 24.84 | 13.68 | −2.1 | 24.31 | 90.87 |
CSR_SR | 25,642 | 15 | 5.38 | −2.07 | 15 | 28.19 |
CSR_ER | 25,642 | 2.48 | 2.68 | 0 | 1.89 | 21.15 |
CSR_SCRR | 25,642 | 1.33 | 3.92 | 0 | 0 | 20 |
CSR_EnvR | 25,642 | 1.37 | 4.19 | 0 | 0 | 30 |
CSR_SocR | 25,642 | 4.7 | 3.55 | −5 | 4.7 | 30 |
CSR_P | 25,642 | 4.77 | 3.49 | −8.64 | 4.77 | 10 |
CSR_S | 25,642 | 1.66 | 0.3 | 1.11 | 1.66 | 2.65 |
CSR_R | 25,642 | 3.42 | 2.33 | 0 | 3.42 | 8 |
CSR_ID | 25,642 | 5 | 0 | 5 | 5 | 5 |
SR_I | 25,642 | 0.25 | 0.62 | 0 | 0 | 4 |
CSR_Perf | 25,642 | 1.79 | 1.18 | 0 | 1.79 | 21.15 |
CSR_Sft | 25,642 | 0.38 | 1.13 | 0 | 0 | 5 |
CSR_EC | 25,642 | 0.31 | 0.96 | 0 | 0 | 5 |
CSR_PQ | 25,642 | 0.64 | 1.88 | 0 | 0 | 9 |
CSR_AS | 25,642 | 0.24 | 0.76 | 0 | 0 | 4 |
CSR_IR | 25,642 | 0.45 | 1.36 | 0 | 0 | 7 |
CSR_EG | 25,642 | 1.37 | 4.19 | 0 | 0 | 30 |
CSR_CV | 25,642 | 4.7 | 3.55 | −5 | 4.7 | 30 |
(a) | ||||
Method | MSE | RMSE | MAE | R2 |
Pooled OLS | 0.010 | 0.098 | 0.066 | 0.076 |
Fixed Effects | 0.009 | 0.095 | 0.064 | 0.107 |
Random Effects | 0.009 | 0.096 | 0.065 | 0.095 |
System GMM | 0.008 | 0.092 | 0.062 | 0.131 |
XGBoost | 0.007 | 0.082 | 0.055 | 0.299 |
(b) | ||||
Model | MSE | RMSE | MAE | R2 |
Feature set 1 | ||||
Decision Tree Regressor | 0.009 | 0.093 | 0.063 | 0.100 |
KNeighbors Regressor | 0.009 | 0.094 | 0.062 | 0.093 |
Ridge Regressor | 0.009 | 0.095 | 0.064 | 0.073 |
XGBoost Regressor | 0.008 | 0.087 | 0.059 | 0.215 |
Random Forest Regressor | 0.008 | 0.089 | 0.060 | 0.185 |
LightGBM Regressor | 0.008 | 0.089 | 0.060 | 0.182 |
CatBoost Regressor | 0.008 | 0.092 | 0.062 | 0.130 |
SVM Regressor | 0.010 | 0.100 | 0.081 | 0.027 |
AdaBoost Regressor | 0.010 | 0.100 | 0.079 | 0.023 |
Feature set 2 | ||||
Decision Tree Regressor | 0.008 | 0.092 | 0.061 | 0.131 |
KNeighbors Regressor | 0.010 | 0.098 | 0.065 | 0.001 |
Ridge Regressor | 0.009 | 0.095 | 0.064 | 0.076 |
XGBoost Regressor | 0.006 | 0.080 | 0.054 | 0.340 |
Random Forest Regressor | 0.009 | 0.094 | 0.063 | 0.086 |
LightGBM Regressor | 0.007 | 0.086 | 0.058 | 0.233 |
CatBoost Regressor | 0.008 | 0.090 | 0.061 | 0.158 |
SVM Regressor | 0.011 | 0.103 | 0.084 | 0.104 |
AdaBoost Regressor | 0.011 | 0.106 | 0.088 | 0.165 |
Feature set 3 | ||||
Decision Tree Regressor | 0.009 | 0.092 | 0.062 | 0.121 |
KNeighbors Regressor | 0.010 | 0.098 | 0.065 | 0.007 |
Ridge Regressor | 0.009 | 0.095 | 0.064 | 0.077 |
XGBoost Regressor | 0.007 | 0.082 | 0.055 | 0.299 |
Random Forest Regressor | 0.008 | 0.091 | 0.061 | 0.149 |
LightGBM Regressor | 0.007 | 0.085 | 0.058 | 0.252 |
CatBoost Regressor | 0.008 | 0.089 | 0.060 | 0.175 |
SVM Regressor | 0.011 | 0.103 | 0.084 | 0.104 |
AdaBoost Regressor | 0.009 | 0.097 | 0.075 | 0.019 |
Feature | Pre-COVID Mean SHAP | COVID Mean SHAP | Δ Mean SHAP | t-Statistic | p-Value | Effect Size (Cohen’s d) |
---|---|---|---|---|---|---|
Top1 | 0.040 | 0.029 | −0.011 | 5.234 | <0.001 *** | 0.612 |
Firmsize | 0.039 | 0.030 | −0.009 | 4.872 | <0.001 *** | 0.523 |
SOE | 0.038 | 0.028 | −0.010 | 5.103 | <0.001 *** | 0.547 |
Lnhp | 0.036 | 0.025 | −0.011 | 5.412 | <0.001 *** | 0.584 |
Mshare | 0.034 | 0.023 | −0.011 | 5.326 | <0.001 *** | 0.569 |
Marketization | 0.033 | 0.024 | −0.009 | 4.687 | <0.001 *** | 0.498 |
MA | 0.032 | 0.021 | −0.011 | 5.234 | <0.001 *** | 0.558 |
SA | 0.031 | 0.022 | −0.009 | 4.523 | <0.001 *** | 0.483 |
CSR Total | 0.030 | 0.024 | −0.006 | 2.892 | 0.004 ** | 0.312 |
Media | 0.028 | 0.020 | −0.008 | 3.926 | <0.001 *** | 0.421 |
Lottery | 0.027 | 0.026 | −0.001 | 0.523 | 0.601 | 0.058 |
HHI | 0.026 | 0.025 | −0.001 | 0.487 | 0.626 | 0.052 |
EPU | 0.025 | 0.008 | −0.017 | 7.234 | <0.001 *** | 0.823 |
Customer | 0.024 | 0.020 | −0.004 | 2.103 | 0.036 * | 0.226 |
Carbontrade | 0.023 | 0.025 | +0.002 | −1.082 | 0.279 | −0.116 |
Multi | 0.022 | 0.018 | −0.004 | 2.234 | 0.026 * | 0.241 |
PLD_DUM | 0.021 | 0.017 | −0.004 | 2.347 | 0.019 * | 0.253 |
Minwage | 0.020 | 0.015 | −0.005 | 2.687 | 0.007 ** | 0.289 |
Digital | 0.019 | 0.014 | −0.005 | 2.523 | 0.012 * | 0.271 |
DIF | 0.018 | 0.011 | −0.007 | 3.412 | <0.001 *** | 0.367 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Wang, Y.; Wei, W.; Liu, Z.; Liu, J.; Lv, Y.; Li, X. Interpretable Machine Learning Framework for Corporate Financialization Prediction: A SHAP-Based Analysis of High-Dimensional Data. Mathematics 2025, 13, 2526. https://doi.org/10.3390/math13152526
Wang Y, Wei W, Liu Z, Liu J, Lv Y, Li X. Interpretable Machine Learning Framework for Corporate Financialization Prediction: A SHAP-Based Analysis of High-Dimensional Data. Mathematics. 2025; 13(15):2526. https://doi.org/10.3390/math13152526
Chicago/Turabian StyleWang, Yanhe, Wei Wei, Zhuodong Liu, Jiahe Liu, Yinzhen Lv, and Xiangyu Li. 2025. "Interpretable Machine Learning Framework for Corporate Financialization Prediction: A SHAP-Based Analysis of High-Dimensional Data" Mathematics 13, no. 15: 2526. https://doi.org/10.3390/math13152526
APA StyleWang, Y., Wei, W., Liu, Z., Liu, J., Lv, Y., & Li, X. (2025). Interpretable Machine Learning Framework for Corporate Financialization Prediction: A SHAP-Based Analysis of High-Dimensional Data. Mathematics, 13(15), 2526. https://doi.org/10.3390/math13152526