Skip to main content
. 2020 Apr 7;10:6026. doi: 10.1038/s41598-020-63159-5

Figure 1.

Figure 1

DeepMicro framework. An autoencoder is trained to map the input X to the low-dimensional latent space with the encoder and to reconstruct X with the decoder. The encoder part is reused to produce a latent representation of any new input X that is in turn fed into a classification algorithm to determine whether the input is the positive or negative class.