A Hybrid Fuzzy-XGBoost Framework for Predicting Business Economic Performance and Assessing Financial Stability

Abstract

Assessing financial stability and enhancing business economic performance are crucial for sustainable growth, sound decision-making, and increased shareholder value in today's unstable economic climate, where businesses face an increasing risk of financial distress and bankruptcy due to unfavourable internal and external factors. In order to reduce redundant, noisy, and irrelevant features while successfully addressing class imbalance and enhancing data quality, this study suggests a machine learning-based framework that combines data normalisation, SMOTE oversampling, and feature selection. Using the enterprise financial network dataset, a fuzzy XGB classifier is created to assess business economic performance and financial stability. Standard performance criteria, including as accuracy, precision, recall, and F1-score, are used to assess the suggested Fuzzy XGB framework. The Fuzzy XGB classifier has a 95.08% classification accuracy, 93.48% precision, 91.76% recall, and an F1-score of 95.93%, according to experimental results, demonstrating its superior predictive ability and classification efficacy. The results verify that a strong and dependable framework for evaluating financial stability and improving business economic performance is provided by the combination of sophisticated data preprocessing methods, feature selection, and the Fuzzy XGB classifier. This framework facilitates the early detection of financially vulnerable organisations and supports strategic financial planning and sustainable organisational growth in dynamic economic conditions.