Quadratic exponential smoothing method is a method to exponentially smooth the first exponential smoothing value again. It can't predict independently, so we must establish a mathematical model to predict with the first exponential smoothing method, and then use the mathematical model to determine the predicted value. The two limiting factors of the first moving average method only exist in the linear second moving average method. The linear quadratic exponential sum smoothing method can be calculated by using only three data and an α value. In most cases, the linear quadratic exponential smoothing method is usually preferred as the prediction method.