Overview

Random Neural Network Simulator implemented in Python.

Setup

Requirements

• Python 3.6+
• NumPy
• Sklearn

Installation

Install this library directly into an activated virtual environment:

$ pip install rnnsim

or add it to your Poetry project:

$ poetry add rnnsim

Usage

After installation, the package can either be used as:

from rnnsim.model import SequentialRNN

sequential_model = SequentialRNN([2, 2, 1])
sequential_model.compile()
sequential_model.fit(train_data=(X_train, y_train), epochs=50, metrics="acc")
print(sequential_model.score((X_test, y_test)))

or

from rnnsim.RNN import RNN

# define model connections
conn_plus = {
    1: [3, 4], 2: [3, 4],
    3: [5], 4: [5], 5: []}
conn_minus = {
    1: [3, 4], 2: [3, 4],
    3: [5], 4: [5], 5: []}
model = RNN(n_total=5, input_neurons=2, output_neurons=1, conn_plus=conn_plus, conn_minus=conn_minus)
model.fit(epochs=N_Iterations, train_data=(X, Y))

References

1. E. Gelenbe, Random neural networks with negative and positive signals and product form solution," Neural Computation, vol. 1, no. 4, pp. 502-511, 1989.
2. E. Gelenbe, Stability of the random neural network model," Neural Computation, vol. 2, no. 2, pp. 239-247, 1990.