Comparison of Feedforward and Recurrent Neural Networks for Predicting Pavement Roughness Why do we need a recurrent neural network model? Key Takeaways:
Comparison of Feedforward and Recurrent Neural Networks for Predicting Pavement Roughness Why do we need a recurrent neural network model? Key Takeaways:
复制标题
用于预测路面粗糙度的前馈神经网络和循环神经网络的比较 为什么我们需要循环神经网络模型？
DOI：
10.1109/bigdata52589.2021.9671404
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Key Takeaways
中科院分区：
文献类型：
--
作者：
Key Takeaways
·虽然RNN模型可能需要更长的时间进行训练，但随着产生更多更好的数据，它在路面使用性能预测方面具有巨大的潜力。基于绩效的规划(PBP)是缓解交通机构普遍面临的预算不足问题的重要工具。实施PBP的一个关键因素是对未来路面状况的有效预测。这取决于稳健的恶化预测模型。
• While the RNN model may take longer to train, it has significant potential for pavement performance prediction as more and better data are generated. Performance-based planning (PBP) is an important tool to mitigate the pervasive problem of inadequate budgets faced by transportation agencies. A key element for implementing PBP is efficient prediction of future pavement conditions. This depends on a robust deterioration prediction model.