This repository is implementation of neural network from the scratch in C++. There are 3 codes. The first code is a neural network without hidden layer which is called Perceptron and is able to predict a class type based on an input data when the classes are only linearly separable. In other words, it acts like linear regression.
The second file (NN_with_Hidden_Layer.cpp) is implementation of a neural network with a single hidden layer which is able to predict a class type based on an input data when the separation between the classes is not possible using a straight line and the relationship between the data is complex and nonlinear.
The third code is a real application of neural network for prediction of an eye disease based on input data-set (IIH_Prediction_Neural_Network.cpp). The data set has been saved in a text file (EYE-DATA.txt).
For more information, please visit my website: https://jafararashmehr.github.io/Computational-BioMechanics/rst/ANN.html
