TSpice-Neuro - he unique features of the Spice-MLP (Multi-Layer Neural Network) are that it allows users choose among 27 activated functions in both hidden and output layers, and hidden - output layers can also use different activated functions.
Graph of Weights and Average Inputs of a Neuron are recorded, this helps users know and understand the convergence of training process.
Data can be split for training and testing. Weights of a trained Network can be saved in binary or text files that help users understand inside the network.
Modeling data is saved in csv format that is easy to export on other application. It has friendly GUI with English, Japanese and Vietnamese interfaces.