Fig. 7: The CNN structure with intermediate layer visualization. | npj Computational Materials

Fig. 7: The CNN structure with intermediate layer visualization.

From: Data-driven analysis of process, structure, and properties of additively manufactured Inconel 718 thin walls

Fig. 7

The CNN architecture consists of three convolutional layers with max pooling layer and the ReLU activation, and fully connected layers with the ReLU activation. The top three subplots are the visualization of output of the first, second, and third convolutional layers. The inputs of 1D CNN are the pre-processed thermal histories. The output is mechanical properties (UTS).

Back to article page