Short-Term IoT Data Forecast of Urban Public Bicycle Based on the DBSCAN-TCN Model for Social Governance
Table 4
Specific implementation steps of the DBSCAN-TCN model.
Step
Content
Step 1
The bike-sharing data is collated into a suitable input format. The data is fed into the model according to chronological order.
Step 2
The structure and parameters of the model are determined. For example, the dilation factor, the number of network layers, the size of the convolutional kernel, the number of layers in the hidden layers, and the number of neurons per layer.
Step 3
The learning rate, the appropriate optimization technique (Dropout), and the appropriate activation function (ReLu) are all selected.
Step 4
The training dataset is used to train the optimization model. The parameters of the prediction model are trained.
Step 5
The validation dataset is used to verify the model predictions. If the prediction is good, the model goes to Step 6; otherwise, it returns to Step 2.
Step 6
The model with the best prediction and the test dataset are used to predict the number of shared bicycle rentals.