Predicting Non Performing Loan of Business Bank with Data Mining Techniques

Wan Jie,Yue Zeng-lei,Yang Dong-hui,ZhangYu,Liu Jiao,Liu Zhi,Liu Jinfu 보안공학연구지원센터 2016 International Journal of Database Theory and Appli Vol.9 No.12

The non-performing loans (NPL) prediction plays an important role in business bank. However, there is still a large gap between the requirement of prediction performance and current techniques. In this paper data mining approaches is used to predict the NPL. Both macroeconomic and bank-specific variables are collected to form the feature set firstly. Based on selected features, the study firstly applies single basic classifiers such as decision tree, k nearest neighbors and support vector machine (SVM) to model the problem of NPL. Bagging and AdaBoost are described in this paper as two different method of multiple classifier fusion, to build prediction models. In this experiment, non-performing loans data with 96 features and 10415 instances of a business bank is collected. F-mean and The Area under the ROC Curve (AUC) are considered as metrics of classification. The results illustrate that multiple classifier fusion algorithms outperform single basic classifier. The model built by multiple classifiers fusion can produce better prediction results. Furthermore, the AdaBoost method performs much better than bagging method in processing NPL.

Ginsenoside Rg1 supplementation clears senescence-associated β-galactosidase in exercising human skeletal muscle

Jinfu Wu,Suchada Saovieng,I-Shiung Cheng,Tiemin Liu,Shangyu Hong,Chang-Yu Lin,I-Chen Su,Chih-Yang Huang,Chia-Hua Kuo 고려인삼학회 2019 Journal of Ginseng Research Vol.43 No.4

Background: Ginsenoside Rg1 has been shown to clear senescence-associated beta-galactosidase (SA-β-gal) in cultured cells. It remains unknown whether Rg1 can influence SA-β-gal in exercising human skeletal muscle. Methods: To examine SA-β-gal change, 12 young men (age 21±0.2 years) were enrolled in a randomized double-blind placebo controlled crossover study, under two occasions: placebo (PLA) and Rg1 (5 mg) supplementations 1 h prior to a high-intensity cycling (70% VO2max). Muscle samples were collected by multiple biopsies before and after cycling exercise (0 h and 3 h). To avoid potential effect of muscle biopsy on performance assessment, cycling time to exhaustion test (80% VO2max) was conducted on another 12 participants (age 23±0.5 years) with the same experimental design. Results: No changes of SA-β-gal were observed after cycling in the PLA trial. On the contrary, nine of the 12 participants showed complete elimination of SA-β-gal in exercised muscle after cycling in the Rg1 trial (p < 0.05). Increases in apoptotic DNA fragmentation (PLA: +87% vs. Rg1: +133%, p < 0.05) and CD68+(PLA:+78% vs. Rg1:+121%, p=0.17) occurred immediately after cycling in both trials. During the 3-h recovery, reverses in apoptotic nuclei content (PLA:+5% vs. Rg1:-32%, p<0.01) and increases in inducible nitrate oxide synthase and interleukin 6 mRNA levels of exercised muscle were observed only in the Rg1 trial (p < 0.01). Conclusion: Rg1 supplementation effectively eliminates senescent cells in exercising human skeletal muscle and improves high-intensity endurance performance.

Ginsenoside Rg1 supplementation clears senescence-associated β-galactosidase in exercising human skeletal muscle

Wu, Jinfu,Saovieng, Suchada,Cheng, I-Shiung,Liu, Tiemin,Hong, Shangyu,Lin, Chang-Yu,Su, I-Chen,Huang, Chih-Yang,Kuo, Chia-Hua The Korean Society of Ginseng 2019 Journal of Ginseng Research Vol.43 No.4

Background: Ginsenoside Rg1 has been shown to clear senescence-associated beta-galactosidase (SA-${\beta}$-gal) in cultured cells. It remains unknown whether Rg1 can influence SA-${\beta}$-gal in exercising human skeletal muscle. Methods: To examine SA-${\beta}$-gal change, 12 young men (age $21{\pm}0.2years$) were enrolled in a randomized double-blind placebo controlled crossover study, under two occasions: placebo (PLA) and Rg1 (5 mg) supplementations 1 h prior to a high-intensity cycling (70% $VO_{2max}$). Muscle samples were collected by multiple biopsies before and after cycling exercise (0 h and 3 h). To avoid potential effect of muscle biopsy on performance assessment, cycling time to exhaustion test (80% $VO_{2max}$) was conducted on another 12 participants (age $23{\pm}0.5years$) with the same experimental design. Results: No changes of SA-${\beta}$-gal were observed after cycling in the PLA trial. On the contrary, nine of the 12 participants showed complete elimination of SA-${\beta}$-gal in exercised muscle after cycling in the Rg1 trial (p < 0.05). Increases in apoptotic DNA fragmentation (PLA: +87% vs. Rg1: +133%, p < 0.05) and $CD68^+$ (PLA:+78% vs. Rg1:+121%, p = 0.17) occurred immediately after cycling in both trials. During the 3-h recovery, reverses in apoptotic nuclei content (PLA:+5% vs. Rg1 -32%, p < 0.01) and increases in inducible nitrate oxide synthase and interleukin 6 mRNA levels of exercised muscle were observed only in the Rg1 trial (p < 0.01). Conclusion: Rg1 supplementation effectively eliminates senescent cells in exercising human skeletal muscle and improves high-intensity endurance performance.

Low Level Segmentation of Motion Capture Data based on Hierarchical Clustering with Cosine Distance

Yang Yang,Jinfu Chen,Zhanzhan Liu,Yongzhao Zhan,Xinyu Wang 보안공학연구지원센터 2015 International Journal of Database Theory and Appli Vol.8 No.4

3D motion capture is to track and record human movements. In recent years, it has been applied into many fields, such as human computer interaction, animation, etc. Low-level segmentation of motion capture data is of significance to the various applications of 3D motion capture; however, due to the high dimensionality of motion capture data, traditional low-level segmentation methods can hardly work out a suitable segmentation for motion capture data. In order to solve this problem, a low-level temporal segmentation algorithm based on cosine distance is proposed, hierarchical clustering is explored so that similar velocity vectors are clustered together according to the cosine distance in a progressive way, the center of each cluster is updated as the vector derived with linear regression, the segment boundaries are determined as the point when the cosine distance between adjacent velocity vectors is greater than 1 (angle>90 degrees). We have conducted experiments on the motion capture database provided by Carnegie Mellon University (CMU), the experiment results show that the performance of the proposed method is optimistic.

Effects of Momordica charantia Saponins on In vitro Ruminal Fermentation and Microbial Population

Kang, Jinhe,Zeng, Bo,Tang, Shaoxun,Wang, Min,Han, Xuefeng,Zhou, Chuanshe,Yan, Qiongxian,He, Zhixiong,Liu, Jinfu,Tan, Zhiliang Asian Australasian Association of Animal Productio 2016 Animal Bioscience Vol.29 No.4

This study was conducted to investigate the effects of Momordica charantia saponin (MCS) on ruminal fermentation of maize stover and abundance of selected microbial populations in vitro. Five levels of MCS supplements (0, 0.01, 0.06, 0.30, 0.60 mg/mL) were tested. The pH, $NH_3-N$, and volatile fatty acid were measured at 6, 24, 48 h of in vitro mixed incubation fluids, whilst the selected microbial populations were determined at 6 and 24 h. The high dose of MCS increased the initial fractional rate of degradation at t-value = 0 ($FRD_0$) and the fractional rate of gas production (k), but decreased the theoretical maximum of gas production ($V_F$) and the half-life ($t_{0.5}$) compared with the control. The $NH_3-N$ concentration reached the lowest concentration with 0.01 mg MCS/mL at 6 h. The MSC inclusion increased (p<0.001) the molar proportion of butyrate, isovalerate at 24 h and 48 h, and the molar proportion of acetate at 24 h, but then decreased (p<0.05) them at 48 h. The molar proportion of valerate was increased (p<0.05) at 24 h. The acetate to propionate ratio (A/P; linear, p<0.01) was increased at 24 h, but reached the least value at the level of 0.30 mg/mL MCS. The MCS inclusion decreased (p<0.05) the molar proportion of propionate at 24 h and then increased it at 48 h. The concentration of total volatile fatty acid was decreased (p<0.001) at 24 h, but reached the greatest concentration at the level of 0.01 mg/mL and the least concentration at the level of 0.60 mg/mL. The relative abundance of Ruminococcus albus was increased at 6 h and 24 h, and the relative abundance of Fibrobacter succinogenes was the lowest (p<0.05) at 0.60 mg/mL at 6 h and 24 h. The relative abundance of Butyrivibrio fibrisolvens and fungus reached the greatest value (p<0.05) at low doses of MCS inclusion and the least value (p<0.05) at 0.60 mg/mL at 24 h. The present results demonstrates that a high level of MCS quickly inhibits in vitro fermentation of maize stover, while MCS at low doses has the ability to modulate the ruminal fermentation pattern by regulating the number of functional rumen microbes including cellulolytic bacteria and fungi populations, and may have potential as a feed additive applied in the diets of ruminants.