ABSTRACT

This study uses the Random Forest algorithm, one of the most popular machine learning methods, to predict the presence of marine species based on environmental factors. The research data was collected from the Cape May Whale Watch and Research Center in Cape May, New Jersey, which includes marine megafauna observations from 2012 to 2024. The data consists of both text and numerical information, all of which were analyzed and cleaned using Python libraries. Additionally, the data was encoded to ensure the Random Forest model could be trained effectively. The results of this study aim to identify which environmental factors are most important for predicting the presence of large marine animals, as shown in the feature importance chart. Model performance was also visualized using graphs of accuracy and loss. Future research will expand the dataset by exploring additional factors and testing alternative machine learning methods to further improve prediction accuracy.