MODELING OF THE BANK CLIENTS CREDITABILITY ASSESSMENT BY A ARTIFICIAL INTELLIGENCE METHOD

Abstract

The article examines the existing methods of credit assessing, their shortcomings and advantages. An analysis of approaches to assessing the bank client’s creditworthiness is proved the need to develop a model, which implementation help to analyze all the bank clients and determine the lower limit of creditworthiness, which will be a normative indicator, considering the existing conditions. The basic method for the model that should be the most common and simplest is a scoring estimate. A tool that allows to analyze a large amount of information is the method of data mining. Based on the analysis of shortcomings of existing methods, requirements for the model of credit rating of bank clients are formulated. The method of modeling the bank's creditworthiness assessment with the help of artificial intelligence method is proposed. Approach consists in the identification and analysis of a large number of unstructured information using statistical and mathematical methods in order to obtain new, structured data on the creditworthiness of the bank's clients. The method of using the artificial intelligence method is in this sequence of actions: 1. determination of output quantities; 2. converting categorical data into numeric; 3. implementation of the forecast; 4. estimation of the obtained model using Fisher's coefficient, R-square and Student's coefficient; 5. checking the significance of each factors by constructing the sum of squares matrix and vector products, calculating the standard deviation for each coefficients, t-statistics and p-value of the error. The model is implemented on the real data and its advantages are defined, which consists in the ability to teach the existing model to determine the lower limit of credit based on really existing customer characteristics, which significantly increases the efficiency of determining the bank's insolvent clients.

Authors and Affiliations

D. V. Bilenko, I. G. Syvytska, D. G. Telenkova

Keywords

Related Articles

USE OF PRIVATE-LABELING IN THE FIELD OF SERVICES

Own trademark or private label – a brand of product or service created by a certain company on request and presented under the brand of another company. Particularly, the practice of using their own TM is widespread in t...

INFORMATION NATURE OF STRUCTURAL CREATIVITY COMPOSITION OF ECONOMIC SYSTEM

Researching of development and efficiency functioning of the economy in the conditions of the formation of an information society actualized the problem of studying the informational essence of the economic system and fa...

PROBLEMS OF THE ENERGY EFFICIENCY MEASURES’ IMPLEMENTATION AT INDUSTRIAL ENTERPRISES OF UKRAINE THROUGH BANKING LOANS

The article deals with the basics of energy efficient measures and implementation of key adverse factors in industrial enterprises of Ukraine. Research is devoted to the solution of the problem of irrational and ineffic...

SOCIOLOGICAL REVIEWS OF THE ENTERPRISE EMPLOYEES AS AN INSTRUMENT FOR THE EMPLOYER RATING

The article analyzes the correlation between sociological reviews of enterprise employees and the employer rating. An HR branding system will allow the enterprise to become a good employer with good businesses, and ensur...

THE FUNCTIONAL CHARACTERISTICS OF CENTERS OF THE PROFESSIONAL RESPONSIBILITY IN THE AUDIT OF FINANCIAL INFORMATION

The research carried out in the article has provided an opportunity to identify and characterize the centers of professional responsibility of the auditor that is one of the its element. The Center of professional respo...

Download PDF file
  • EP ID EP562134
  • DOI -
  • Views 137
  • Downloads 0

How To Cite

D. V. Bilenko, I. G. Syvytska, D. G. Telenkova (2018). MODELING OF THE BANK CLIENTS CREDITABILITY ASSESSMENT BY A ARTIFICIAL INTELLIGENCE METHOD. Проблеми системного підходу в економіці, 3(65), -. https://www.europub.co.uk/articles/-A-562134