Credit Risk Analysis of Taiwan’s Financial Sector with Data Mining Method

摘要

With explosive growth in both information and data mining technology, the possession of large databasesrncreates ample business opportunity. With this phenomena ever so prevalent nowadays, it is challenging tornscour through useful data to aid the financial sector’s decision-making process. Thus, this scholarship uses thernfinancial sector’s vast database to analyze credit card transaction data, data mining algorithms that applyrnconsumer clustering, as well as use such clustering results to classify algorithms that contribute to customerrnclustering. To adhere to customer classification, data mining models are used to find problematic householdsrnto precisely predict a high-risk group customer base. As a result, this precision provides a foundation forrnfinancial institutions to practice risk diversification and management. For customers without potential risk,rnsuch institutions can thus practice better deals, bonuses and diversified packages to increase customer loyalty.rnSimply put, such institutions’ ultimate goals are to hold a position in the customer group’s top tier.