A Python implementation of Neural Network without hidden layers. You can use it to train a neural network, and then use it to predict for new samples.
neurons.py provides the class to solve the problem. It follows the
func.py is function library. Currently, it implements
relu and their derivatives.
gen.py is used to generate the data. It takes in a configuration json file and several command line arguments. The directory used to store the data should already exist. Otherwise, it will reports error, saying can't find files or directories.
demo.ipynb shows how to use this package.
All rights reserved © 2017 Wenjie Zheng
Please contact me if you need more information about this package.