How to Develop 1D Convolutional Neural Network Models for Human Activity Recognition

The updated version of the evaluate_model() function is listed below that creates a three-headed CNN model. We can see that each head of the model is the same structure, although the kernel size is varied. The three heads then feed into a single merge layer before being interpreted prior to making a prediction. When the model is created, a plot of the network architecture is created; provided below, it gives a clear idea of how the constructed model fits together.