Internet of Things Enabled Financial Crisis Prediction in Enterprises Using Optimal Feature Subset Selection-Based...

Area of Science:

• Business Analytics
• Computational Finance
• Machine Learning

Background:

• Accurate forecasting of business failure and financial crises is crucial for small- to medium-sized enterprises (SMEs).
• Existing prediction models may lack the precision required for timely intervention.
• The integration of advanced computational techniques offers potential for improved financial risk assessment.

Purpose of the Study:

• To develop and validate an optimal feature selection (FS)-based classification model for financial crisis prediction (FCP) in SMEs.
• To enhance the accuracy and reliability of financial distress forecasting.
• To provide an effective tool for early detection of potential business failures.

Main Methods:

• Data acquisition using Internet of Things (IoT) devices.
• Pigeon-Inspired Optimization (PIO) for optimal feature selection.
• Extreme Gradient Boosting (XGB) classification optimized by Jaya Optimization (JO) algorithm (JO-XGB).

Main Results:

• The proposed PIO-JO-XGBoost model demonstrated superior performance in financial crisis prediction.
• Experimental validation confirmed the model's effectiveness compared to existing methods.
• The feature selection and optimized classification significantly improved prediction accuracy.

Conclusions:

• The developed PIO-JO-XGBoost model is an effective tool for financial crisis prediction in SMEs.
• The combination of PIO for FS and JO for XGBoost optimization enhances predictive capabilities.
• This approach offers a robust solution for mitigating financial risks in businesses.