Data Mapping System Of Riau Province Fire Potential Using K-Means Clustering Method

Abstract

According to a report from the Riau Province BLHK states that hotspots in Riau Province are always present every year despite the number of hotspots that have been suppressed (http://dislhk.riau.go.id/). One of the causes is the frequent land clearing occurred as a trigger from a hotspot in Riau Province. There is a need for countermeasures as soon as possible to overcome the problem of hotspots that will cause forest fires. These problems need to be watched out quickly, one of which is to know in advance the hotspots that are likely to emerge based on existing data. Data mining processing is very suitable to be applied in order to produce relevant data to find out the possibility of hotspots. In this study the data grouping was done in the form of a visualization of hotspot mapping using the K-means Clustering method. The parameters used include 3 number of clusters (critical, alert, vigilant), 12 regencies / cities in Riau Province and 3 attributes (hotspots, number of fires, number of events). With the results of the visualization of the mapping using the K-means Clustering method, it is expected to be able to help the relevant parties, namely the Riau Provincial Forest Service in handling early the hotspots that are likely to emerge.

Analyzing the Effect of Similarity Functions on Partitioning Around Medoids Algorithm for Mapping DHF Disease in North Sumatra

“…Clustering is one of the tasks mining that can be used for mapping based on similar characteristics [2] [2]. Partitioning Around Medoid (PAM) is a well-known clustering algorithm for mapping various cases.…”