aics_code/2_11.py

import tensorflow as tf
import numpy as np

# prepare dataset
def raw_data():
    for i in range(2**4):
        for j in range(2**4):
            yield [[float(x) for x in '%04d%04d'%(int(bin(i)[2:]),int(bin(j)[2:]))], [float(x) for x in '%05d'%int(bin(i + j)[2:])]]
raw = list(raw_data())
rng = np.random.default_rng()
rng.shuffle(raw)
raw_len = len(raw)

train_len = int(raw_len * 0.8)
train_data, train_label = zip(*raw[:train_len])
val_data, val_label = zip(*raw[train_len:])

# learning params
learning_rate = 0.1
training_epochs = 200
batch_size = 16

# network params
n_hidden_1 = 30
n_hidden_2 = 11
n_input = 8
n_output = 5

# model
x = tf.placeholder("float", [None, n_input])
y = tf.placeholder("float", [None, n_output])

def mlp(_X, _weights, _biases):
    layer1 = tf.nn.sigmoid(tf.add(tf.matmul(_X, _weights['h1']), _biases['b1']))
    layer2 = tf.nn.sigmoid(tf.add(tf.matmul(layer1, _weights['h2']), _biases['b2']))
    return tf.matmul(layer2, _weights['out']) + _biases['out']

weights = {
    'h1' : tf.Variable(tf.random_normal([n_input, n_hidden_1])),
    'h2' : tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
    'out' : tf.Variable(tf.random_normal([n_hidden_2, n_output]))
}

biases = {
    'b1' : tf.Variable(tf.random_normal([n_hidden_1])),
    'b2' : tf.Variable(tf.random_normal([n_hidden_2])),
    'out' : tf.Variable(tf.random_normal([n_output]))
}

pred = mlp(x, weights, biases)
cost = tf.reduce_mean(tf.square(pred - y))
optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(cost)

init = tf.global_variables_initializer()

train_data_batches = np.array_split(train_data, len(train_data) // batch_size)
train_label_batches = np.array_split(train_label, len(train_label) // batch_size)

correct_prediction = tf.equal(y, tf.where(pred<0.5, x = tf.zeros_like(pred), y = tf.ones_like(pred)))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))

# training

with tf.Session() as sess:
    sess.run(init)
    for epoch in range(training_epochs):
        for i, j in zip(train_data_batches, train_label_batches):
            sess.run(optimizer, feed_dict={x: i, y: j})
        print("epoch:", epoch, "Train Accuracy:", accuracy.eval({x: train_data, y: train_label}), "Val Accuracy:", accuracy.eval({x: val_data, y: val_label}))