Individual demand forecasting based on fuzzy Markov chain model with weights

According to the randomness and self-correlation of individual demand, it is discussed in the necessity and feasibility of the introduction of fuzzy Markov chain model with weights to predict the future individual demand. The specific steps are explained: set up the classification by the standard deviation of sales series, and weighted by the standardized self-coefficients, calculated the transition probability matrix and the state probability. Then, a concrete forecasting value was obtained by using the level characteristics value of fuzzy sets. An example is presented on the sales forecasting of fast moving consumer goods in instant customerization, and it showed that the fuzzy Markov chain model with weights (FMCW) is more suitable for individual demand forecasting, compared with the Moving Average, Simple Exponential Smoothing, Linear Regression, and GM (1,1).