Interpretable decision trees through MaxSAT

Alos, Josep; Ansótegui, Carlos; Torres, Eduard

doi:10.1007/s10462-022-10377-0

Cited by 4 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The former refers to the design of AI models with their interpretability in mind. For example, transparent models such as decision trees (Alòs et al, 2023; Florez‐Lopez & Ramon‐Jeronimo, 2015; Niu, Zhang, et al, 2020), generalized additive models (Chen et al, 2018), rule‐based models (Yang, Lim, et al, 2023), and causality (Shin, 2021), among others. Yang, Lim, et al (2023) proposed an integrated learning approach based on evidential inference rules that provide the contribution of attributes and activation rules for a good understanding of model predictions.…”

Section: Review Of the Literaturementioning

confidence: 99%

Credit risk prediction based on causal machine learning: Bayesian network learning, default inference, and interpretation

Liu,

Zhang,

Xiong

2024

Journal of Forecasting

View full text Add to dashboard Cite

The predictive and interpretable power of models is crucial for financial risk management. The purpose of this study was to perform credit risk prediction in a structured causal network with four stages—data processing, structural learning, parameter learning, and interpretation of inferences—and use six real credit datasets to conduct empirical research on the proposed model. Compared with traditional machine learning algorithms, we comprehensively explain the results of credit default through forward and reverse reasoning. We also compared our model with the post hoc interpretation models local interpretable model‐agnostic explanations (LIME) and shapley additive explanations (SHAP) to verify the interpretability of Bayesian networks. The experimental results show that the prediction performance of Bayesian networks is superior to traditional machine learning models and similar to the performance of ensemble models. Furthermore, Bayesian networks offer valuable insights into the interplay of features by considering their causal relationships and enable an assessment of how individual features influence the prediction outcome. In this study, what‐if analysis was performed to assess credit default probabilities under various conditions. This analysis provides decision‐makers with the necessary tools to make informed judgments about the risk profile of borrowers. Consequently, we consider Bayesian networks as a viable tool for credit risk prediction models in terms of prediction performance and interpretability.

show abstract

Section: Review Of the Literaturementioning

confidence: 99%

Credit risk prediction based on causal machine learning: Bayesian network learning, default inference, and interpretation

Liu,

Zhang,

Xiong

2024

Journal of Forecasting

View full text Add to dashboard Cite

show abstract

“…In [30], the authors presented a hybrid classifier system that used clustering to improve the accuracy of the decision tree. In [31], the authors provided a method for enhancing machine learning decision trees' accuracy-interpretability trade-off. In [32], the authors used confusion matrix to optimize decision tree.…”

Section: Data Classificationmentioning

confidence: 99%

AI-Based Employment Situation Analysis of Innovative Talents in the New Media Era

Ma¹

2023

J Multimed Inf Syst

View full text Add to dashboard Cite

As the number of undergraduates increases, competition for employment opportunities becomes more intense. To improve employment guidance and increase the employment rate and quality, it is essential to evaluate preliminary employment data, identify the primary variables that influence employment, and provide decision support for decision-makers. In the era of new media, the influx of massive information can make it difficult for undergraduates to distinguish between true and false information, making it necessary to analyze the employment situation of innovative talents using artificial intelligence. However, historical employment data is generally a dataset with high noise due to the diversity and complexity of actual employment situations. Differences in accuracy requirements among various institutions also make handling issues such as different accuracy requirements and noise adaptation abilities challenging. To address this issue, this study proposes applying a decision tree algorithm based on a multi-scale rough set model to analyze university employment data. By introducing the multiscale concept into rough set theory and drawing on variable precision rough sets, the analysis's accuracy and noise adaptation ability can be improved. Experimental results show that the proposed algorithm performs well in terms of classification accuracy and running time, and the employment situation of creative minds can be effectively analyzed. Moreover, the decision tree has a simple structure, concise rules, no inseparable datasets, and a fast running speed.

show abstract

Reservoir fluid identification based on multi-head attention with UMAP

Hua,

Gao,

et al. 2024

Geoenergy Science and Engineering

View full text Add to dashboard Cite

Interpretable decision trees through MaxSAT

Cited by 4 publications

References 20 publications

Credit risk prediction based on causal machine learning: Bayesian network learning, default inference, and interpretation

Credit risk prediction based on causal machine learning: Bayesian network learning, default inference, and interpretation

AI-Based Employment Situation Analysis of Innovative Talents in the New Media Era

Reservoir fluid identification based on multi-head attention with UMAP

Contact Info

Product

Resources

About