An Improved Grey Time-Delay Multivariable Model and Its Application

An accurate output value prediction is of great significance for policy-making and plan development in high-tech industries. An adaptive cumulative time-delay discrete grey multivariable model is proposed to forecast the output value of high-tech industries in China. Specifically, a new time-delay function is constructed to unify the time-delay effects and facilitate a realistic adaptation to disparate timedelay effects. Subsequently, the Grey Wolf Optimizer algorithm is used to find the optimal parameters for the new time-delay function, which improves the model's adaptability. Furthermore, to further enhance the model's prediction accuracy, an iterative reweighted least squares method is also adopted to optimize the parameter vector. To demonstrate the applicability of the proposed model, it is employed in hightech industries in China. The empirical findings show that the proposed model outperforms the benchmark models.