论文部分内容阅读
提出了一个基于单整自回归移动平均模型(ARIMA)与神经网络模型(PSONN)的成本竞争优势预测模型。将成本竞争优势序列分解成线性结构序列以及非线性结构残差序列,前者用ARIMA模型来预测,残差部分用PSONN模型来进行非线性预测。预测结果表明,组合模型的预测精度显著高于单个线性或非线性模型的预测结果,同时结果也表明了中国制造业在未来八年内仍然保有成本竞争优势,但是来自生产效率方面的贡献在降低,提示制造业需从单纯强调劳动力成本的发展模式过渡到以提高劳动生产率为重心的道路上,采用“适度增长”的劳动力成本与“高速增长”的劳动生产率相结合的发展模式。
A cost competitive advantage prediction model based on single-autoregressive moving average model (ARIMA) and neural network model (PSONN) is proposed. The cost competitive advantage sequence is decomposed into linear structural sequence and non-linear structural residual sequence. The former is predicted by ARIMA model and the other is predicted by PSONN model. The prediction results show that the prediction accuracy of the combined model is significantly higher than that of a single linear or nonlinear model. The results also show that China’s manufacturing industry still retains the cost competitive advantage within the next eight years, but the contribution from the production efficiency is declining. It is suggested that the manufacturing industry should adopt a mode of development that combines the labor cost of “modest growth ” with the labor productivity of “high-speed growth ” on the transition from the mode of development emphasizing labor cost alone to the focus of improving labor productivity.