Keras

from keras import Sequential
from keras.layers import Dense
import random
import numpy as np


model = Sequential()
model.add(Dense(32, activation='relu', input_dim=100))
model.add(Dense(1, activation='sigmoid'))
model.compile(optimizer='rmsprop',
              loss='binary_crossentropy',
              metrics=['accuracy'])

def get_random_power(power, samples=100, interval=(0.5,2)):
    m = random.random() * (interval[1]-interval[0]) + interval[0]
    return [x**power * m for x in range(samples)]

def np_get_random_power(power, samples=100, interval=(0.5,2)):
    m = random.random() * (interval[1]-interval[0]) + interval[0]
    return np.array(x**power * m for x in range(samples))

def sinus(offset, samples = 100):
    import math
    return [math.sin(offset*x/samples) for x in range(samples)]

def get_data_set(size):
    X = []
    Y = []
    choice = [0,1]
    for i in range(size):
        val = random.choice([0,1])
        if val == 0:
            X.append(get_random_power(val))
        else:
            X.append(sinus(val,))
        Y.append([val])
    return X,Y

X,Y = get_data_set(1000)
print(np.array(X).shape, np.array(Y).shape)

# Generate dummy data
data = np.random.random((1000, 100))
print("Shape:", data.shape)
labels = np.random.randint(2, size=(1000, 1))
print("Shape:", labels.shape)

# Train the model, iterating on the data in batches of 32 samples
model.fit(np.array(X), np.array(Y), epochs=10, batch_size=32)

X,Y = get_data_set(1000)
result = model.predict(np.array(X))
for i in range(len(Y)):
    print(int(result[i]*1000)/1000,Y[i])