Advanced Financial Stress Prediction Using Random Forest and Logistic Regression: A Machine Learning Approach

This research aims to develop a model that utilizes technology to assess the level of financial hardship in companies in Indonesia. Over the period 2013 to 2023, 324 companies from various sectors (excluding the banking sector) were analyzed. The method applied is random forest, a machine learning technique that divides data into training and testing datasets and is able to handle large and imbalanced data. To support this research, random forest utilizes big data analytics and cloud computing technologies to process data efficiently. These technologies allow the random forest to make decisions quickly based on the data received. In addition, logistic regression is applied to determine the variables that affect the prediction of financial distress. The research shows that the random forest model has high accuracy, with five financial ratio variables affecting financial distress prediction. This research opens opportunities for the development of machine learning-based applications to predict financial trouble in companies in Indonesia. By simply inputting financial ratio data, companies can quickly obtain prediction results. This technology allows companies to address financial distress earlier and more efficiently, improving their ability to survive in dynamic economic conditions.