Predicting Business Performance With Ai: a Quantitative Comparison of Traditional Vs. Ai-driven Forecasting Models Among Msmes in Klang Valley
Year: 2026
Type: article
Abstract: This study presents a comprehensive quantitative evaluation of artificial intelligence (AI) driven forecasting models compared to traditional statistical methods in predicting business performance among Micro, Small, and Medium Enterprises (MSMEs) in Malaysia's Klang Valley. Through analysis of 180 MSMEs across manufacturing, services, and retail sectors over a 36-month period, this research examines the predictive accuracy, reliability, and practical implementation of various forecasting approaches. The study employs advanced machine learning algorithms including Random Forest, Neural Networks, and Support Vector Machines alongside traditional methods such as ARIMA, Exponential Smoothing, and Linear Regression. Performance evaluation utilizes multiple accuracy metrics including Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). Results demonstrate that AI-driven models achieve superior predictive accuracy with average RMSE improvements of 23.7% and MAE reductions of 31.2% compared to traditional methods. The research provides empirical evidence supporting AI adoption for business forecasting while identifying implementation challenges and success factors specific to MSMEs in developing economies. Keywords: AI forecasting; business performance prediction; MSMEs; machine learning; traditional forecasting
Source: Jurnal Apresiasi Ekonomi
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7203734009