Hands-On Machine Learning for Algorithmic Trading

How to build on a pre-trained CNN

The transfer learning approach to CNN relies on pre-training on a very large dataset such as ImageNet. The goal is that the convolutional filters extract a feature representation that generalizes to new images. In a second step, it leverages the result to either initialize and retrain a new CNN or as inputs to in a new network that tackles the task of interest.

As discussed, CNN architectures typically use a sequence of convolutional layers to detect hierarchical patterns, adding one or more fully-connected layers to map the convolutional activations to the outcome classes or values. The output of the last convolutional layer that feeds into the fully-connected part is called bottleneck features. We can ...