Abstract
Due to the enormous increasing number of burning cases, Jabatan Bomba dan Penyelamat Malaysia(JBPM) has been given responsibility to deal with burning cases in Malaysia. According to some research, the causes of burning cases include firecrackers, short circuit (electricity) , ciggrattes, spark of fire mosquito coil, kitchen gases,spontaneous reaction, intentional burning, unknown sources,chemical reaction, and matches. The forecast of number of burning cases in Johor has been carried out for the purpose to help JBPM, Larkin Johor in dealing and managing those cases in more sistematically way. The purpose of this project is to analyze and forecast the number of burning cases in Johor Darul Takzim based on sources that frequently causes burning. Three models for the furpose to forecast the number of burning cases within next 12 months have been developed using neural network method. That is, the sources of burning are intentional burning, unknown sources and electricity. The software used to develop this model is the 7th version software package of Matlab in Windows. The analysis and forecasting are made based on monthly data for the past 13 years starting from 1993 to 2005.
Website : http://ir.fsksm.utm.my/1069/
No comments:
Post a Comment