Clustering Lakes in Kampar by Using Morphometry Data and Machine Learning Analysis

Abstract

Lakes have an important role in hydrological and biochemical cycle. It also has some other crucial role such as domestic and industrial water use as well as irigation. The monitoring and management of this aquatic resources is crucial. But with many numbers of lakes, it is very challenging to manage them all. Clustering lakes can provide the answer so the management of the same cluster lakes may be done efficiently. Within this study, morphometry data of 6 lakes in Kampar Regency, Riau Province, were analyzed by using one of artificial intelligence branch which is machine learning. Morphometrical data are collected by using information geographic system. These data then categorized by using python language. This categorization based on data mining categorization algorithm named K-means. Based on the K-means machine learning clustering, the optimum cluster based on Elbow methid is k=3. But there is a possibility to look around for k=2. Based on K=3, cluster 3 is defined as the lowest values of all atribute. Based on k=2, the lowest value of morphometry data wiil be in the cluster 1.. These data will not only provide basic data such as total area, shape, width and length, but also help to understand the large scale hydrological models.