Research on Personal Credit Risk Assessment Model Based on Instance-Based Transfer Learning

Personal credit risk assessment is an important part of the development of financial enterprises. Big data credit investigation is an inevitable trend of personal credit risk assessment, but some data are missing and the amount of data is small, so it is difficult to train. At the same time, for different financial platforms, we need to use different models to train according to the characteristics of the current samples, which is time-consuming. In view of these two problems, this paper uses the idea of transfer learning to build a transferable personal credit risk model based on Instance-based Transfer Learning (Instance-based TL). The model balances the weight of the samples in the source domain, and migrates the existing large dataset samples to the target domain of small samples, and finds out the commonness between them. At the same time, we have done a lot of experiments on the selection of base learners, including traditional machine learning algorithms and ensemble learning algorithms, such as decision tree, logistic regression, xgboost and so on. The datasets are from P2P platform and bank, the results show that the AUC value of Instance-based TL is 24% higher than that of the traditional machine learning model, which fully proves that the model in this paper has good application value. The model’s evaluation uses AUC, prediction, recall, F1. These criteria prove that this model has good application value from many aspects. At present, we are trying to apply this model to more fields to improve the robustness and applicability of the model;on the other hand, we are trying to do more in-depth research on domain adaptation to enrich the model.

AHP-FCE Based Physical Exercise Risk Evaluation Model

Exercise is a highly proven and beneficial health promotion modality, But it is very difficult to determine whether the person during exercise is safe. A unique and comprehensive approach is proposed to perform physical exercise risk evaluation （PERE）, in which personalized factors are deterrrdned basing on grey correlation analysis, analytic hierarchy process （AHP） method is used to structure the large numbers of risk factors, and fuzzy comprehensive evaluation （FCE） is applied to fuzzify the factors and compute the exercise risk level. Finally, an actual calculation example is used to verify the feasibility of the method.

Validation of different personalized risk models of chemotherapy-induced nausea and vomiting:results of a randomized,double-blind,phase III trial of fosaprepitant for cancer patients treated with high-dose cisplatin

Background:Highly emetogenic chemotherapy induces emesis in cancer patients without prophylaxis.The purpose of this study was to evaluate the efficacy and safety of a fosaprepitant-based triple antiemetic regimen for the prevention of chemotherapy-induced nausea and vomiting(CINV)in patients with solid malignant tumors,determine risk factors and externally validate different personalized risk models for CINV.Methods:This phase III trial was designed to test the non-inferiority of fosaprepitant toward aprepitant in cancer patients who were to receive the first cycle of single-day cisplatin chemotherapy.The primary endpoint was complete response(CR)during the overall phase(OP)with a non-inferiority margin of 10.0%.Logistic regression modelswere used to assess the risk factors ofCRand no nausea.To validate the personalized risk models,the accuracy of the risk scoring systems was determined by measuring the specificity,sensitivity and area under the receiver operating characteristic(ROC)curve(AUC),while the predictive accuracy of the nomogram was measured using concordance index(C-index).Results:A total of 720 patients were randomly assigned.CR during the OP in the fosaprepitant group was not inferior to that in the aprepitant group(78.1%vs.77.7%,P=0.765)with a between-group difference of 0.4%(95%CI,-5.7%to 6.6%).Female sex,higher cisplatin dose(≥70 mg/m2),no history of drinking and larger body surface area(BSA)were significantly associated with nausea.The AUC for the acute and delayed CINV risk indexes was 0.68(95%CI:0.66-0.71)and 0.66(95%CI:0.61-0.70),respectively,and the C-index for nomogram CINV prediction was 0.59(95%CI,0.54-0.64).Using appropriate cutoff points,the three models could stratify patients with high-or low-risk CINV.No nausea and CR rate were significantly higher in the low-risk group than in the high-risk group(P<0.001).Conclusions:Fosaprepitant-based triple prophylaxis demonstrated non-inferior control for preventing CINV in patients treated with cisplatin-base chemotherapy.Female cancer patients without a history of alcohol consumption,with larger BSA and received high-dose cisplatin might be more vulnerable to CINV.Three personalized prediction models were well-validated and could be used to optimize antiemetic therapy for individual patients.

The application of omics technologies to toxicology

The application of omics technologies,including genomics,transcriptomics,proteomics and metabolomics,has the potential to revolutionize toxicology by providing a more comprehensive understanding of the molecular mechanisms of toxicity,identifying potential biomarkers of exposure or effect,and enabling personalized risk assessments for individuals.Each omics approach has its own challenges,including data analysis and interpretation,but the integration of multiple omics approaches can provide a more comprehensive understanding of toxicity.The use of omics technologies for personalized risk assessments can inform targeted interventions and improve public health outcomes.While challenges remain,the potential benefits of omics technologies in toxicology make it an exciting area of research for the future.