import numpy as np
X = np.array(([2,9], [1,5], [3,6]), dtype = float)
y = np.array(([.92], [.86], [.89]), dtype = float)
X = X/np.amax(X, axis = 0)

def sigmoid(x):
    return 1/(1+np.exp(-x))

def der_sigmoid(x):
    return x*(1-x)

epoch = 5000
lr = 0.01
neurons_i = 2
neurons_h = 3
neurons_o = 1
suni 
weight_h = np.random.uniform(size=(neurons_i, neurons_h))
bias_h = np.random.uniform(size=(1, neurons_h))
weight_o = np.random.uniform(size=(neurons_h, neurons_o))
bias_o = np.random.uniform(size=(1, neurons_o))

for i in range(epoch):
    inp_h = np.dot(X, weight_h) + bias_h
    out_h = sigmoid(inp_h)
    
    inp_o = np.dot(out_h, weight_o) + bias_o
    out_o = sigmoid(inp_o)
    
    err_o = y - out_o
    grad_o = der_sigmoid(out_o)
    delta_o = err_o * grad_o
    
    err_h = delta_o.dot(weight_o.T)
    grad_h = der_sigmoid(out_h)
    delta_h = err_h * grad_h
    
    weight_o += out_h.T.dot(delta_o) * lr
    weight_h += X.T.dot(delta_h) * lr
    
print('Input:n', X)
print('Actual:n', y)
print('Predicted:n', out_o)