Implementation of data mining on mangroves species using k-means algorithm

K-means clustering helps to find the hidden pattern in the data behavior and extract insights from it. It is a process of grouping data objects based on their variations and similarities. It is often used in the field of Biology. Silo-siloan Islet is a Marine Protected Area (MPA) with various mangrove species. It is a marine environment reserved by the local government of Clarin, Bohol, to protect its natural resources. The source of the dataset was based on the assessment results conducted on the identification of mangroves on the islet. It was found that there were two species of mangroves, namely: Rhizophora mucronata and Sonneratia alba. Each mangrove tree was measured according to its height, crown cover, crown diameter, and Girth at breast height (GBH). The k-means algorithm was used to classify the mangrove species according to their similar characteristics. At the same time, the Elbow method was also used to find the optimum number of clusters. It was found that there were two possible clusters in the. The result has a precision of 0.80 for cluster 0 and 0.97 for cluster 1. The level of recall for cluster 0 is 0.98 and 0.86 for cluster 1; meanwhile, both of the clusters have an f1-score level of 0.81. Thus, the overall weighted mean in precision is 0.89, 0.92, and 0.81 for recall and f1-score, which means the clustering model is reliable. It was found that all Rhizophora mucronata shared uniform characteristics in terms of cover, diameter, GBH, and height. In comparison, the majority of Sonneratia alba shared the same characteristics as Rhizophora mucronata. The results of this study will be used as input for the local government to take action on the importance of mangroves forest.