EVALUATION OF PERSONAL CREDITABILITY ON THE BASIS OF ARTIFICIAL INTELLIGENCE METHODS
Journal Title: Вісник Кременчуцького національного університету імені Михайла Остроградського - Year 2017, Vol 1, Issue 107
Abstract
Purpose. To investigate the possibilities of improving existing methods for determining the assessment of the personal creditworthiness of potential borrowers based on classification algorithms for further usage in the field of personal credit banking systems. Methodology. When solving the problem, general methods of data mining have been used, such as the decision tree, to identify potential borrowers, as well as existing approaches to machine learning to assess personal credit based on the probabilistic Naïve Bayes classifier. Findings. In this paper, the issue of determining personal creditworthiness has been considered, the existing approaches and in the issues of determining personal creditworthiness have been analyzed. The analysis of the personal data of potential borrowers has been carried out to reveal the possibilities of reducing the dimension of many factors included in the credit model. The developed model has been investigated by the Weka software application, which made it possible to visualize the results of identifying potential borrowers. Originality. Today the definition of personal credit evaluation in the field of personal banking credit systems does not have one best solving method. The article presents the results of our own research in this field, which can be used in the development of general methods of estimating personal creditworthiness. Practical value. This paper shows the practical possibility of reducing the dimension of the classification model without reducing its accuracy, which makes it possible to reduce the cost of constructing a classification tree and a probabilistic table in the process of machine learning.
Authors and Affiliations
M. Smirnova, Hu Xiaohui, A. Judina
Keywords
Related Articles
- EP ID EP660624
- DOI -
- Views 66
- Downloads 0
SOFTWARE-HARDWARE COMPLEX FOR THE TASK OF PERSONAL VOICE AUTHENTICATION
Purpose. During the pre-study mathematical models of human voice signal as a localized human voice structures has been developed. To approbate and fulfill mathematical models of human voice for the task of personal authe...
MULTI-OBJECTIVE OPTIMIZATION OF STEADY OPERATING MODES OF ELECTRIC OIL PUMPING STATION
Purpose. To research multi-objective optimization of steady operating modes of electric oil pumping station basing on system approach by criteria of efficiency and reliability of electric and hydraulic subsystems. Method...
DISTRIBUTION OF PUBLIC PROCUREMENT AS AN INSTRUMENT OF FINANCIAL POLICY IN HIGHER EDUCATION
Purpose. In the article, as a panacea for chronic underfunding of state higher education, it is used a conceptual method, that allows the state to simultaneously both the regulator of relations in the field of higher edu...
IMPROVING ORGANIZATIONAL STRUCTURE OF INNOVATIION MANAGEMENT AT MACHINE-BUILDING ENTERPRISES
Purpose. The paper focuses on studying the structure of the enterprise's innovation goals, finding out impact factors in the structure of innovation management, developing options for the organizational structure with th...
JUSTIFICATION OF FEASIBILITY OF MEDICAL WASTE RECYCLING AND PROCESSING BY INCINERATION AT THE HEALTHCARE FACILITIES
The purpose. To implement the installation for medical waste utilization in medical and preventive measures. Methodology. Performance of an analysis for medical waste generation volume in administrative districts of Kyi...