Densely connected neural networks for nonlinear regression
This work addresses regression challenges in environmental data analysis, though it is incremental as it adapts an existing architecture to a new task.
The authors tackled the problem of information loss in convolutional DenseNets for regression tasks by proposing a fully connected DenseNet model, which achieved a high correlation of 0.91 in predicting relative humidity and outperformed baseline models like support vector regression.
Densely connected convolutional networks (DenseNet) behave well in image processing. However, for regression tasks, convolutional DenseNet may lose essential information from independent input features. To tackle this issue, we propose a novel DenseNet regression model where convolution and pooling layers are replaced by fully connected layers and the original concatenation shortcuts are maintained to reuse the feature. To investigate the effects of depth and input dimension of proposed model, careful validations are performed by extensive numerical simulation. The results give an optimal depth (19) and recommend a limited input dimension (under 200). Furthermore, compared with the baseline models including support vector regression, decision tree regression, and residual regression, our proposed model with the optimal depth performs best. Ultimately, DenseNet regression is applied to predict relative humidity, and the outcome shows a high correlation (0.91) with observations, which indicates that our model could advance environmental data analysis.