Efficacy of Machine Learning Models in Predicting Ocean pH Levels

Abstract

This paper will investigate the efficacy of machine learning models, including Linear Regressor, MLP regressor, Support Vector Machine (SVM), and Random Forest Regressor, in accurately predicting ocean pH levels. By utilizing a comprehensive dataset of oceanographic variables, we evaluate the models' performance on training and development sets. The findings highlight the relative unimportance of eight oceanographic variables in the prediction of ocean pH levels using the machine learning models mentioned above. These insights contribute to a better understanding of ocean acidification impacts and will aid in the development of mitigation strategies because scientists will be able to focus their efforts on the three important variables given by the models.