Fire Time Series Forecasting Based on Markov-SVR Model

Based on support vector regression and Markovstate transition, a new prediction model termed as Markovsupportvector regression (MSVR) model is proposed toforecast the fire time series. In the proposed model, a SVR is tobuild an optimal prediction model from a series of fire data,and then uses the Markov state transition to reduce theresiduals errors produced by the mentioned model. Theproposed model is examined using actual fire time series data.The results show that the MSVR model gets the better resultperformance than that of the pure SVR model.