Marine diesel engines are crucial for powering large vessels in the maritime sector and are known for their efficiency across various industries. However, increasing environmental concerns and stringent regulations targeting air pollutants such as nitrogen oxides (NOx) and particulate matter (PM) have heightened the need for advanced emission control technologies. Addressing this challenge, the study focuses on developing a reliable method to predict NOx emission levels in marine engines, reducing reliance on resource-intensive experimental testing. Leveraging machine learning techniques, particularly k-nearest neighbors (kNN)-based algorithms, the research classifies NOx emissions in marine engines operating under the Reactivity-Controlled Compression Ignition (RCCI) strategy. Comparative performance analysis reveals that the FPFS-kNN algorithm achieves the highest accuracy (90.00%) alongside strong precision (84.23%), recall (82.37%), and F1 score (82.47%). These findings underscore the potential of machine learning in emission prediction and highlight directions for future exploration in this domain.