Garnik Arakelyan

Mr. Garnik Arakelyan


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
GAGarnik Arakelyan
Garnik Arakelyan
Armen GhazaryanArmen Ghazaryan
Armen Ghazaryan
The banking system, as the most important sector of the economy of every country, often encounters a number of risks. Financial institutions of that system operate in an unstable environment, and without having complete information about that environment, they may suffer significant losses. The main source of such losses is considered to be credit risks, and for the management of these, various mathematical models are being developed which will allow banks to make decisions on granting a loan. Lately, for this purpose, machine learning (ML) classification algorithms have often been used for credit risk modeling. In this research work, using the ideas of well-known ML algorithms, a new algorithm for solving the binary classification problem was developed. By means of the algorithm created, based on real data, a classification model has been developed. Qualitative indicators of that model, such as ROC AUC, PR AUC, precision, recall, and F1 score, were evaluated. By modifying the resulting probabilities into a range of 300–850 score points, a scoring model has been developed, the usage of which can mitigate credit risk and protect financial organizations from major losses.
1

1 January 2022 in Messenger of Armenian State University of Economics
Verified
Journal Article
Closed access
Armen GhazaryanArmen Ghazaryan
Armen Ghazaryan
LGLiana Grigoryan
Liana Grigoryan
GAGarnik Arakelyan
Garnik Arakelyan
There are many problems in each credit institution. The most important of them is the risk of possible losses in lending. Within the framework of the topic, the studies conducted by other researchers were investigated, from which it was concluded that machine learning tools are often used to optimally solve the above-mentioned problem. Real data on credits were used as a basis for modeling in the work. In this work, based on the available data, several machine learning models were developed, from which the best one was selected, which can contribute to the improvement of the credit risk management process. During the work, the logical connections between data and their interaction with each other were revealed. Then, based on the work done, the appropriate models were built, the quality of which was checked using various tools. The obtained models were compared and the best one was selected. The obtained results are practically applicable and show that each bank and credit organization can develop a better solution based on the large databases they have, which will contribute to curbing credit risk and reducing costs.
1

12 August 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
GAGarnik Arakelyan
Garnik Arakelyan
Armen GhazaryanArmen Ghazaryan
Armen Ghazaryan
Banking system being the most important sector of the economy of every country often encounters a number of risks. Financial institutions of that system operate in an unstable environment and without having complete information about that environment, they may suffer significant losses. The main source of such losses is considered to be a credit risk, for the management of which various mathematical models are being developed, which allow to make decision on granting a loan. Lately on this level machine learning (ML) classification algorithms are often used for credit risk modeling. In this research work using the ideas of ML well-known algorithms a new algorithm for solving the binary classification problem was developed. By means of the created algorithm based on real data a classification model has been de-eloped, qualitative indicators of that model such as: ROC AUC, PR AUC, Precision, Re-call, F1 score, were evaluated. By modifying the resulting probabilities into a range of 300-850 score points a scoring model has been developed, the usage of which can mitigate the credit risk and protect financial organizations from major losses.
1

1 July 2024 in Scientific Proceedings of Vanadzor State University: “Natural and Exact Sciences
Verified
Journal Article
Closed access
GAGarnik Arakelyan
Garnik Arakelyan
Key words: logistic regression, kNN, genetic algorithm, mutation, data grouping, correlation Any company, including banks and credit organizations, operates in an unstable environment and may incur significant losses without having complete information about it. One of the main sources of such losses is credit risk, for the management of which various mathematical models are created. However, modeling often faces challenges related to the lack of a sufficient number of observations. Within the research topic, studies by other researchers have been examined. In this work, an attempt was made to create synthetic data based on a small number of real credit observations, which can be used to create machine learning models that require a large dataset. To generate synthetic data, the logic of genetic algorithms, the concepts of Darwin's theory of evolution, as well as machine learning methods that do not require a large amount of data were used. The quality of the generated data was assessed using statistical methods. The results obtained are practically applicable and demonstrate that any bank or credit organization can develop a high-quality solution for managing credit risks even with a small amount of available data.

20 August 2025 in Регион и мир / Region and the World
Verified
Journal Article
Closed access
GAGarnik Arakelyan
Garnik Arakelyan
The banking system is a key sector of the economy and often faces various types of risks. This article focuses on credit risks, which are the main source of financial losses for credit institutions, including banks. Based on real data, a scoring model was developed in this study to assess and optimally manage credit risk, using the XGBoost machine learning algorithm. The model returns score values in the range of 300–850. Quality metrics such as ROC AUC, PR AUC, Precision, Recall, and F1 score were evaluated for the developed model. Within the scope of the research, a new approach was proposed to assess the sensitivity of a machine learning model to changes in variable values. The approach is based on the concept of entropy. The results show that the use of machine learning models makes it possible to assess and optimally manage credit risk, thereby reducing potential losses. Moreover, the proposed new approach, when applied together with the well-known SHAP method, allows for a deeper evaluation of the model and its sensitivity to changes in variable values.