Importance of Consistent and Well-Dispersed Catalyst Datasets for Machine Learning in Oxidative Coupling of Methane

Key Points

• This research aims to investigate the importance of uniform and diverse experimental datasets in catalyst informatics for enhancing machine learning applications.
• Utilized a dataset focused on oxidative coupling of methane
• Data collected by a single researcher for consistency
• Experimental conditions and devices were kept uniform
• Identified that well-dispersed data led to better machine learning outcomes
• Uniform datasets improved model accuracy
• Diversity in datasets contributed to more reliable predictive capabilities

Abstract