This is an efficient implementation of a fully connected neural network in NumPy. The network can be trained by a variety of learning algorithms: backpropagation, resilient backpropagation and scaled conjugate gradient learning. The network has been developed with PYPY in mind.
BSD 2-Clause "Simplified" License
297
stars
98
forks
source link
Print hit rates and print status more frequently #4
Added calculation for total hit rate percentage to resilientbackprop, backprop, and scg methods.
Changed test to return both hit rate percentage and MSE.
Print hit rate percentage, MSE, and epoch on every iteration rather than every 1000.
Added calculation for total hit rate percentage to resilientbackprop, backprop, and scg methods. Changed test to return both hit rate percentage and MSE. Print hit rate percentage, MSE, and epoch on every iteration rather than every 1000.