A prediction method of defaulters of bank loans based on big data mining

摘要

#$%^&*AU2020100708A420200618.pdf#####Abstract This invention is applied to the field of financial data analysis, and the probability model is constructed by combining the observed and recorded data with the actual common sense to determine the possibility of the event. This invention includes the following steps: 1. Make essential preparations, including data collecting, cleaning and importing 2. Build the project analysis process, including TAN, Markov and Markov FS three different models 3. Give the relative importance of prediction components and the relationship between each prediction variable. Evaluate the accuracy of the model. The accuracy of model prediction is verified by comparing the original data with the model prediction data. Finally, we find that Tree Augmented Naive Bayesian model accounts for the highest accuracy rate.Load data Set target quantity -I r Discard the nill value Determine the Bayesian model as the classification model to be used Establish the models under the python IDE Build the TAN Build the Markon Build the Markonmodel model FS model Analyze and predict data under the models The prediction results of the target values of each rnodel are obtained The model is evaluated based on the results Figure I 1