Data mining applications in capital markets: The example of Borsa Istanbul

Data mining is a powerful tool that supports strategic decision-making processes in the financial sector and is used in a wide range of areas, from customer relationship management to risk assessment. Using data mining techniques, financial institutions analyze customer behavior, assess credit risk, and perform important tasks such as fraud detection. In addition, analysis of financial statements can be carried out more effectively using data mining techniques and thus more in-depth information can be obtained about the financial performance of companies. In this study, it is aimed to evaluate the financial health and risk level of companies traded in Borsa İstanbul (BİST) using data mining techniques, with data obtained from financial statements such as balance sheets and income statements. For this purpose, first of all, data such as total resources, equity, assets, liabilities and net period profit or loss in the financial statements of 672 BİST companies in various sectors published between 2016 and 2023 were collected through the Public Disclosure Platform (KAP). Then, a data set containing 4048 records was created by calculating the financial risk ratios of the companies. Finally, financial data were classified using Naive Bayes, k Nearest Neighbor (KNN), Decision Tree, Random Forest, Logistic Regression, Deep Learning and Support Vector Machine (SVM) classification algorithms, and the highest success was the Random Forest algorithm with an accuracy value of 98.37%. With this study, companies can predict their future financial situations by using past performance data. In this way, investors and managers can make more informed decisions and minimize financial risks.