Cance level at 0.05; . significance level at 0.1.Dangers 2021, 9,13 ofTable six. Lasso logistic regression
Cance level at 0.05; . significance level at 0.1.Dangers 2021, 9,13 ofTable six. Lasso logistic regression benefits. 2017 Two Years Before Economic Distress Ratios R2 R5 R14 R15 R17 R21 R22 Coefficients 0.0574 -0.0010 10.0928 -7.9388 -0.4502 0.0010 0.0003 2018 1 Year Prior to Economic Distress Ratios R4 R6 R8 R14 R15 R16 R17 R20 R21 Coefficients 0.0937 -0.9277 0.0029 34.9176 -6.5013 -0.0700 -1.2586 -0.1070 0.four.4. Efficiency of Logit models The results obtained by the confusion matrices are determined by the test sample. As shown in Table 7, two years before the occurrence of economic distress, the stepwise logistic regression model appropriately classifies 93.33 from the SMEs. 1 year before the occurrence of economic distress, the (Z)-Semaxanib web accuracy improves to 95.00 along with the sensitivity is 96.67 (29/30 with the failing SMEs are properly classified).Table 7. Confusion matrices for logit models, years: 2017018. Stepwise Logistic Regression Lasso Logistic Regression2017 two years before monetary distress 0 1 0 28 (93.33 ) a 2 (6.67 ) c SC-19220 MedChemExpress Overall accuracy 1 two (6.67 ) b 28(93.33 ) d 93.33 0 1 0 23 (76.67 ) five (16.67 ) General accuracy 1 7 (23.33 ) 25 (83.33 ) 80.002018 a single year before financial distress 0 1 0 28 (93.33 ) 1 (three.33 ) Overall accuracy 1 two (six.67 ) 29 (96.67 ) 95.00 0 1 0 26 (86.67 ) four (13.33 ) All round accuracy 1 four (13.33 ) 26 (86.67 ) 86.67Notes: a indicates the specificity; b indicates the sort II error; c indicates the form I error; d indicates the sensitivity. The rate with the metrics are shown in parentheses. 0 and 1 indicate healthful SMEs and financially distressed SMEs, respectively.Concerning the performance of lasso logistic regression models, the accuracy improves in 2018 with 86.67 in comparison with 80 in 2017. The type I error (When a model classifies a failing organization as healthful) goes from 16.67 in 2017 to 13.33 in 2018 showing the improvement on the top quality in the model when economic distress is imminent. four.five. Efficiency of Neural Networks Models To seek out the most beneficial neural networks models for stepwise logistic selection and lasso logistic choice, we vary the network parameters, namely the hidden layers from 0 to 10 and also the quantity of its nodes from 0 to ten. We come across that the very best neural networks models for stepwise logistic selection (resp for lasso logistic selection) are composed of a single hidden layer containing 3 nodes. In accordance with Table 8, in 2017 the lasso neural networks model performs far better than the stepwise neural networks model with an accuracy of 83.33 . Moreover, the type I error on the lasso neural networks model is 6.67 against 13.33 for the stepwise neural networks model, a difference of six.66 .Dangers 2021, 9,14 ofTable eight. Confusion matrices for neural networks models, years: 2017018. Stepwise Logistic Regression Lasso Logistic Regression2017 two years before financial distress 0 1 0 23 (76.67 ) a four (13.33 ) c General accuracy 1 7 (23.33 ) b 26 (86.67 ) d 81.67 0 1 0 22 (73.33 ) two (6.67 ) General accuracy 1 eight (26.67 ) 28 (93.33 ) 83.332018 one particular year before financial distress 0 1 0 26 (86.67 ) three (10.00 ) General accuracy 1 4 (13.33 ) 27 (90.00 ) 88.33 0 1 0 26 (86.67 ) 4 (13.33 ) All round accuracy 1 four (13.33 ) 26 (86.67 ) 86.67Notes: a indicates the specificity; b indicates the kind II error; c indicates the form I error; d indicates the sensitivity. The price from the metrics are shown in parentheses. 0 and 1 indicate wholesome SMEs and financially distressed SMEs, respectively.As for 2018, the stepwise ne.
