Demonstrating non-inferiority of easy interpretable methods for insolvency prediction

Obermann, Lennart; Waack, Stephan

doi:10.1016/j.eswa.2015.08.009

Cited by 18 publications

(22 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, domain experts cannot understand the mechanism behind the model, so to improve it, further research is necessary. ()…”

Section: Introductionmentioning

confidence: 99%

“…Second, compared to the black box model, both risk control experts in the financial field and lay people are more inclined to trust a model in which the result is understandable. Additionally, machine learning models are likely to be incorrect . Besides, if the mechanism behind the model is understood, domain experts' substantial domain knowledge can be combined with the model, building a more powerful model …”

Section: Introductionmentioning

confidence: 99%

“…Therefore, domain experts cannot understand the mechanism behind the model, so to improve it, further research is necessary. [16][17][18] In other research, the definition of model interpretability is different. The model is interpretable when the importance of the variable importance is specified 19 or the rules are completely understandable, regardless of the number of rules.…”

mentioning

confidence: 99%

“…Additionally, machine learning models are likely to be incorrect. 16 Besides, if the mechanism behind the model is understood, domain experts' substantial domain knowledge can be combined with the model, building a more powerful model. 21 Some scholars have conducted related research on model interpretability, but there are fewer studies on interpretability compared with model accuracy.…”

mentioning

confidence: 99%

See 3 more Smart Citations

A hybrid interpretable credit card users default prediction model based on RIPPER

Ding

Pan

2018

Concurrency and Computation

View full text Add to dashboard Cite

Summary With the vigorous development of the financial sector, financial risks are showing a tendency toward diversification, particularly regarding the customer credit risk of commercial banks. Therefore, the customer's credit risk is being considered by financial institutions, and a credit evaluating model has emerged as a result. Currently, research has concentrated on enhancing the precision of the model, ignoring the interpretability, which makes it difficult to apply in the industry. Compared to precision, studies related to the interpretable model are limited. In our previous work, we did not consider model operation time and stability. Therefore, this study proposes a hybrid model based on the RIPPER algorithm. First, according to the characteristics of credit card data sets, targeted special data pretreatment methods are proposed. Next, the RELIEF method for feature selection removes the redundant features and further improves the interpretability of the model. Then, to address the problem of the imbalanced distribution of credit card data sets, a synthetic minority class sampling algorithm is used to equalize the samples. Finally, default credit card users are predicted by taking advantage of the rules generated by the RIPPER algorithm. To test the performance of the model, we used Taiwanese credit card customer data for empirical research. We considered model accuracy and interpretability when comparing the proposed SPR‐RIPPER model with the existing mainstream models. The results of the experiments indicate that the proposed model achieves acceptable results. This study demonstrates that the proposed credit card user default prediction model, SPR‐RIPPER, has practical application value.

show abstract

“…Therefore, domain experts cannot understand the mechanism behind the model, so to improve it, further research is necessary. ()…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

A hybrid interpretable credit card users default prediction model based on RIPPER

Ding

Pan

2018

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…Prediction of insolvency and rating predictions are essential in the world of finances. Inaccurate predictions may result in erroneous decisions and as a result lead to huge financial losses (Obermann & Waack, 2015). For this reason in practice systems forecasting the financial situation of enterprises are applied (Early Warning Systems -EWS).…”

Section: Introductionmentioning

confidence: 99%

Assessment applicability of selected models of multiple discriminant analyses to forecast financial situation of Polish wood sector enterprises

Adamowicz

Noga

2017

Folia Forestalia Polonica

View full text Add to dashboard Cite

In the last three decades forecasting bankruptcy of enterprises has been an important and difficult problem, used as an impulse for many research projects (Ribeiro et al. 2012). At present many methods of bankruptcy prediction are available. In view of the specific character of economic activity in individual sectors, specialised methods adapted to a given branch of industry are being used increasingly often. For this reason an important scientific problem is related with the indication of an appropriate model or group of models to prepare forecasts for a given branch of industry. Thus research has been conducted to select an appropriate model of Multiple Discriminant Analysis (MDA), best adapted to forecasting changes in the wood industry. This study analyses 10 prediction models popular in Poland. Effectiveness of the model proposed by Jagiełło, developed for all industrial enterprises, may be labelled accidental. That model is not adapted to predict financial changes in wood sector companies in Poland.The generally known Altman model showed the greatest effectiveness in the identification of enterprises at risk of bankruptcy. However, that model was burdened with one of the greatest errors in the classification of healthy enterprises as sick. The best effectiveness in the identification of enterprises not threatened with bankruptcy was found for forecasts prepared using the Prusak 2 model. However, forecasts based on those models were characterised by erroneous classification of sick companies as healthy. The model best fit to predict the financial situation of Polish wood sector companies was the Poznań model Pz = 3.562 · X 1 + 1.588 · X 2 + 4.288 · X 3 + 6.719 · X 4 ) -2.368 where: X 1 -net income/total assets; X 2 -(current assets -stock)/current liabilities; X 3 -fixed capital/total assets X 4 -income from sales/sales revenue).

show abstract

Interpretable speech features vs. DNN embeddings: What to use in the automatic assessment of Parkinson’s disease in multi-lingual scenarios

Favaro,

Tsai,

Butala

et al. 2023

Computers in Biology and Medicine

View full text Add to dashboard Cite

Demonstrating non-inferiority of easy interpretable methods for insolvency prediction

Cited by 18 publications

References 59 publications

A hybrid interpretable credit card users default prediction model based on RIPPER

A hybrid interpretable credit card users default prediction model based on RIPPER

Assessment applicability of selected models of multiple discriminant analyses to forecast financial situation of Polish wood sector enterprises

Interpretable speech features vs. DNN embeddings: What to use in the automatic assessment of Parkinson’s disease in multi-lingual scenarios

Contact Info

Product

Resources

About