Search

This project focuses on automating the credit card application assessment process using advanced machine learning techniques, including Random Forest, Gradient Boosting, SVMs, Logistic Regression, Regularization Methods, and Hyperparameter Tuning. The objective is to improve the efficiency, accuracy, and fairness of credit card approval decisions. Historical credit card application data, comprising applicant demographics, financial history, and employment details, is collected and pre-processed. Feature engineering and exploratory data analysis (EDA) enhance the dataset’s predictive power. Three machine learning algorithms, Random Forest, Logistic Regression, and Gradient Boosting are applied. Regularization techniques (L1 and L2) and hyperparameter tuning are used to prevent overfitting and optimize model performance. The project assesses model performance by employing metrics such as accuracy, precision, recall, F1-score, and ROC-AUC metrics, and conducts feature importance analysis to identify key factors influencing approval decisions. The project aims to deliver robust, accurate, and fair credit card approval models, benefitting both financial institutions and applicants

The paper discusses the evaluation of the financial stability of an enterprise, depending on the business potential of a municipality. The aim of the research is to prove a relation of the prediction models and their ability to predict the financial ability of an enterprise depending on sample clusters of the business potential in a municipality. In the first part of the research, the municipalities are clustered using the cluster analysis and the municipalities are classified into four clusters - economic, social, infrastructural, and environmental. Subsequently, financial data on business entities located in the area are used and the prediction models are calculated using the Czech indices of the Neumaiers. The results are further graphically evaluated in terms of size and sectors of the enterprise. The ANOVA method is verified to assess the independence of the creditworthiness models on sample clusters. Regarding the method and the analytical tool, it is possible to state their suitable choice; the outputs are relevant, valid, and usable in practice. The influence of the creditworthiness on the environmental cluster in the area is proved.

The uncertainty in the financial market is often perceived as a risk of deviation from expected results. However, uncertainty is associated with vagueness in the sense of ambiguity or obscurity, which, unlike the risk, is not describable in the form of deterministic or stochastic models. Given the vagueness of the data entering the credit risk assessment models, the fuzzy decision-making process is introduced in the theoretical part and implemented in the application part as an effective alternative to standard models. The fuzzy approach is used to address the problem of the creditworthiness of the corporate bonds issuers and the credit risk. Data sources are of a qualitative and quantitative nature based on the ratings of Standard and Poor's and Moody's and financial indicators relevant for the creditor. The resulting creditworthiness and credit risk values rectify the credit ratings of rating agencies and are the original contribution to the topic solved. In conclusion, the impact of company ownership on the creditor´s subjective perception of creditworthiness, or respectively credit risk is outlined.

In this research, analyzing the difficulties of efficiency of commercial banks’ credit operations, the author has paid special attention to such issues like the process of management and assessment of borrowers’ creditworthiness, as well as formation of credit portfolio in unsteady conditions of economics. The main risk of formation of commercial bank’s credit portfolio is made up of the credit risk. In the present situation in Latvia, when during the pre-crisis years the commercial banks dramatically increase the volumes of their credit portfolios, the basic challenge for them is balance of profitability of credit portfolio and quality. Here, if the mechanisms of the market self-regulation do not function, only efficient supervising system can help. Latvia is one of the first countries of the Eastern Europe, which has a united institution of supervision of financial and capital market being established and having commenced its work, which is also a step into the direction of development, safety and stability of the credit market. Analyzing the formation of the credit portfolio of the Latvian commercial banks in the unsteady economical circumstances, researching the statistical data of the Bank of Latvia and the Financial and Capital Market Commission, the author has made conclusions that during the crisis period both the physical volume of credit portfolios o the Latvian commercial banks and their quality and profitability have decreased, together with rapid increase of the volume of overdue credit payments. In this situation the author would recommend paying special attention to more efficient management of the quality of granted credits without materially decreasing the volume of credit portfolios themselves. In its turn, in order to efficiently manage the quality of granted credits, one has to use the results of the local researches, not relying particularly on the foreign researches, which refer to the conjuncture of other markets, specific nature of other economics and other time.