Application of Machine Learning in Medium-Chain Carboxylic Acids Production from Organic Wastes

In addressing global energy demands and waste management challenges, the bioconversion of waste into energy, particularly through medium-chain carboxylic acids (MCCAs) production, represents a promising approach for resource recovery. However, MCCA production faces challenges like incomplete substrate utilization, lack of electron donors, by-product formation, and system instability. Our study introduces machine learning (ML) as a solution to these challenges. By employing operational parameters and genomic data, the ML model achieved a prediction accuracy exceeding 85%, identifying key factors influencing MCCA production rates. Furthermore, the integration of optimization algorithms enabled the model to determine the optimal conditions for MCCA production, significantly improving yields. This approach not only enhances the predictability and optimization of MCCA production but also contributes to more efficient resource management in waste recovery. Our findings represent a significant step forward in sustainable waste-to-energy conversion processes.