基于结构优化的RAN城市环境空气质量预测模型

根据城市空气质量随时间变化的特性,利用资源分配神经网络和隐层节点相关性剪枝方法,建立了一个结构简单、具有在线学习能力的空气质量预测模型。通过对网络模型的训练和测试,表明该模型不仅可降低网络结构的复杂度,而且可以得到比普通资源分配网络更精度的预测结果。

Air quality predicting model based on the resource allocation network of structure optimization

According to the time-varying characteristics of the urban air quality, an air quality predicting model using the resource allocation neural network and the hidden layer nodes correlation pruning algorithm was established. The model had  simple structure and the online learning capability. Training and testing results showed that the model could not only reduce the complexity of the network structure, but obtain more accurate forecasting than the usual resource allocation neural network.