Розробка web-додатку «банківський асистент» з побудовою моделі логістичної регресії

Abstract

The report describes the process of creating a web application serving as a tool for predicting the chance of an approve or reject for being issued a loan. The prediction is based on personal data of an end user (such as age, salary, education, employed years, debt-to-income, etc.) and is implemented via Logistic Regression. The article describes integration and further utilizing Logistic Regression model by the web application. Thus process of an estimation of credit scoring is automated and the outcome (default or not default) is based on historical data. The goal of the web application is to improve the efficiency and accuracy of loan approval.