Basic DNN Example: Approximate sine function using tflearn

tflearn-example-sine-function.py

import tflearn 
import numpy as np
import matplotlib.pyplot as plt

N = 1000

# Approximate sine function using tflearn
x = np.linspace(-np.pi, np.pi, N)
y = np.sin(x) + np.random.uniform(-0.5,0.5, N)

# Model
net = tflearn.input_data(shape=[None,1])
net = tflearn.fully_connected(net, 1)
net = tflearn.fully_connected(net, 32, activation='relu')
net = tflearn.fully_connected(net, 1)
net = tflearn.regression(net, optimizer='adam', loss='mean_square', metric='R2', learning_rate=0.01)

model = tflearn.DNN(net)
model.fit(x.reshape(-1, 1), y.reshape(-1, 1), batch_size=100, n_epoch=100, shuffle=True)

# Test the model
x0 = np.linspace(-np.pi,np.pi,100).reshape(-1, 1)
pred = model.predict(x0)

plt.plot(x0, pred, color="r")
plt.scatter(x, y, alpha=0.1)
plt.show()