Fuzzy Association Rules Mining pada Data Klimatologi dan Jumlah Hotspot di Kalimantan Tengah dan Kalimantan Selatan

Abstract

The occurrence of forest fires was influenced by many factors of weather and climate such as temperature, rainfall, solar radiation, humidity, air stability, speed and direction of wind. Rainfall, relative humidity and temperature were the main factors on the moisture of fuel control. Wind and solar radiation were important factors on drying fuels. It was influenced on change in fuel temperature also temperature and relative humidity in the air. This research was aimed to extract or “mine” knowledge from climatology and number of hotspots data in Central Kalimantan and South Kalimantan by 2001 until 2004. It is using Association Rules Mining method with apriori algorithm to see the relation between the number of hotspots and climatology data. Also fuzzy logic was used to solve an uncertainty and the numerical data. The result of this research was presented in the rules. The rules show relation between climatology and the number hotspots data. In rainy season (November-June) few hotspots were shown. This was influenced by mild temperature, mild solar radiation, mild air pressure, high relative humidity and low wind speed. However in dry season the occurence of forest fires was high in each year. The peak of forest fires occurred in August and September. It was influenced by mild temperatures, low rainfall, low solar radiation, mild air pressure, mild relative humidity, low and mild wind speed