ENN stands for the Evolution of Neural Networks.
The client provided here works by asking a service (source not included here) for a population of neural network architectures in Keras 2.2.4 JSON format.
Upon receving these network descriptions, the client then sends out requests for each of the candidate networks to be trained for a small number of epochs by a (possibly remotely distributed) worker process.
After measurements of each candidate network's performance come back, fitness measurements are sent back to the same ENN service, where the candidate networks are used as a basis to create another new, evolved generation of candidate networks.
Demo below Use Cases: 1.Image Classification i.Chest X-ray Classification ( Multi- Label) ii.UTK Face Age Classification ( Multi- Class) 2.Text Classification i.Toxicity Classification ( Binary) ii.Survey Classification ( Multi- Output)
#Dataset for UTKFace data refer: https://drive.google.com/drive/folders/0BxYys69jI14kU0I1YUQyY1ZDRUE