def nnet(X, Y):
"""Neural net with regularization."""
lambda_ = 1e-4 # Weight regularizer
noise = .5 # Likelihood st. dev.
net = (
ab.InputLayer(name="X", n_samples=1) >>
ab.Dense(output_dim=40, l2_reg=lambda_) >>
ab.Activation(tf.tanh) >>
ab.Dense(output_dim=20, l2_reg=lambda_) >>
ab.Activation(tf.tanh) >>
ab.Dense(output_dim=10, l2_reg=lambda_) >>
ab.Activation(tf.tanh) >>
ab.Dense(output_dim=1, l2_reg=lambda_)
)
f, reg = net(X=X)
lkhood = tf.distributions.Normal(loc=f, scale=noise).log_prob(Y)
loss = ab.max_posterior(lkhood, reg)
return f, loss
def my_model(features, labels, mode, params):
N = params["N"]
n_samples = NSAMPLES if mode == tf.estimator.ModeKeys.TRAIN \
else NPREDICTSAMPLES
X = tf.feature_column.input_layer(features, params['feature_columns'])
kernel = ab.RBF(LENSCALE, learn_lenscale=True)
net = (
ab.InputLayer(name="X", n_samples=n_samples) >>
ab.RandomFourier(n_features=NFEATURES, kernel=kernel) >>
ab.Dense(output_dim=64, init_fn="autonorm") >>
ab.Activation(tf.nn.selu) >>
ab.DenseVariational(output_dim=1, full=False, prior_std=1.0,
learn_prior=True)
)
phi, kl = net(X=X)
std = ab.pos_variable(NOISE, name="noise")
ll_f = tf.distributions.Normal(loc=phi, scale=std)
predict_mean = ab.sample_mean(phi)
# Compute predictions.
if mode == tf.estimator.ModeKeys.PREDICT:
predictions = {
'predictions': predict_mean,
'samples': phi
}
return tf.estimator.EstimatorSpec(mode, predictions=predictions)
ll = ll_f.log_prob(labels)
loss = ab.elbo(ll, kl, N)
tf.summary.scalar('loss', loss)
# Compute evaluation metrics.
mse = tf.metrics.mean_squared_error(labels=labels,
predictions=predict_mean,
name='mse_op')
r2 = r2_metric(labels, predict_mean)
metrics = {'mse': mse,
'r2': r2}
if mode == tf.estimator.ModeKeys.EVAL:
return tf.estimator.EstimatorSpec(
mode, loss=loss, eval_metric_ops=metrics)
# Create training op.
assert mode == tf.estimator.ModeKeys.TRAIN
optimizer = tf.train.AdamOptimizer()
train_op = optimizer.minimize(loss, global_step=tf.train.get_global_step())
return tf.estimator.EstimatorSpec(mode, loss=loss, train_op=train_op)
def nnet_dropout(X, Y):
"""Neural net with dropout."""
lambda_ = 1e-3 # Weight prior
noise = .5 # Likelihood st. dev.
net = (
ab.InputLayer(name="X", n_samples=n_samples_) >>
ab.Dense(output_dim=32, l2_reg=lambda_) >>
ab.Activation(tf.nn.selu) >>
ab.DropOut(keep_prob=0.9, independent=True) >>
ab.Dense(output_dim=16, l2_reg=lambda_) >>
ab.Activation(tf.nn.selu) >>
ab.DropOut(keep_prob=0.95, independent=True) >>
ab.Dense(output_dim=8, l2_reg=lambda_) >>
ab.Activation(tf.nn.selu) >>
ab.Dense(output_dim=1, l2_reg=lambda_)
)
f, reg = net(X=X)
lkhood = tf.distributions.Normal(loc=f, scale=noise).log_prob(Y)
loss = ab.max_posterior(lkhood, reg)
return f, loss
def nnet_ncp(X, Y):
"""Noise contrastive prior network."""
noise = ab.pos_variable(.5)
lstd = 1.
perturb_noise = 10.
net = (
ab.InputLayer(name="X", n_samples=n_samples_) >>
ab.NCPContinuousPerturb(input_noise=perturb_noise) >>
ab.Dense(output_dim=32) >>
ab.Activation(tf.nn.selu) >>
ab.Dense(output_dim=16) >>
ab.Activation(tf.nn.selu) >>
ab.Dense(output_dim=8) >>
ab.Activation(tf.nn.selu) >>
ab.DenseNCP(output_dim=1, prior_std=.1, latent_std=lstd)
)
f, kl = net(X=X)
lkhood = tf.distributions.Normal(loc=f, scale=noise).log_prob(Y)
loss = ab.elbo(lkhood, kl, N)
return f, loss
def linear(X, Y):
"""Linear regression with l2 regularization."""
lambda_ = 1e-4 # Weight regularizer
noise = 1. # Likelihood st. dev.
net = (
ab.InputLayer(name="X") >>
ab.Dense(output_dim=1, l2_reg=lambda_)
)
Xw, reg = net(X=X)
lkhood = tf.distributions.Normal(loc=Xw, scale=noise).log_prob(Y)
loss = ab.max_posterior(lkhood, reg)
# loss = 0.5 * tf.reduce_mean((Y - Xw)**2) + reg
return Xw, loss
def test_categorical_likelihood(make_data, likelihood):
"""Test aboleth with discrete likelihoods.
Since these are kind of corner cases...
"""
x, y, _, = make_data
like, K = likelihood
N, _ = x.shape
# Make two classes (K = 2)
Y = np.zeros(len(y), dtype=np.int32)
Y[y[:, 0] > 0] = 1
if K == 1:
Y = Y[:, np.newaxis]
X_ = tf.placeholder(tf.float32, x.shape)
Y_ = tf.placeholder(tf.int32, Y.shape)
n_samples_ = tf.placeholder(tf.int32)
layers = ab.stack(
ab.InputLayer(name='X', n_samples=n_samples_),
ab.Dense(output_dim=K)
)
nn, reg = layers(X=X_)
like = like(logits=nn)
log_like = like.log_prob(Y_)
prob = like.prob(Y_)
ELBO = ab.elbo(log_like, reg, N)
MAP = ab.max_posterior(log_like, reg)
fd = {X_: x, Y_: Y, n_samples_: 10}
tc = tf.test.TestCase()
with tc.test_session():
tf.global_variables_initializer().run()
assert like.probs.eval(feed_dict=fd).shape == (10, N, K)
assert prob.eval(feed_dict=fd).shape == (10,) + Y.shape
L = ELBO.eval(feed_dict=fd)
L = MAP.eval(feed_dict=fd)
assert np.isscalar(L)
def svr(X, Y):
"""Support vector regressor, kind of..."""
lambda_ = 1e-4
eps = 0.01
lenscale = 1.
# Specify which kernel to approximate with the random Fourier features
kern = ab.RBF(lenscale=lenscale)
net = (
# ab.InputLayer(name="X", n_samples=n_samples_) >>
ab.InputLayer(name="X", n_samples=1) >>
ab.RandomFourier(n_features=50, kernel=kern) >>
# ab.DropOut(keep_prob=0.9, independent=True) >>
ab.Dense(output_dim=1, l2_reg=lambda_)
)
f, reg = net(X=X)
loss = tf.reduce_mean(tf.nn.relu(tf.abs(Y - f) - eps)) + reg
return f, loss
def main():
"""Run the demo."""
# Get Continuous and categorical data
df_train, df_test = fetch_data()
df = pd.concat((df_train, df_test))
X_con, X_cat, n_cats, Y = input_fn(df)
n_samples_ = tf.placeholder_with_default(T_SAMPLES, [])
# Define the continuous layers
con_layer = (
ab.InputLayer(name='con', n_samples=n_samples_) >>
ab.RandomFourier(100, kernel=ab.RBF(learn_lenscale=True)) >>
ab.Dense(output_dim=16, init_fn="autonorm")
)
# Now define the cateogrical layers, which we embed
# Note every Embed call can be different, this is just "lazy"
cat_layer_list = [ab.Embed(EMBED_DIMS, i, init_fn="autonorm")
for i in n_cats]
cat_layer = (
ab.InputLayer(name='cat', n_samples=n_samples_) >>
ab.PerFeature(*cat_layer_list) >> # Assign columns to embedding layers
ab.Activation(tf.nn.selu) >>
ab.Dense(16, init_fn="autonorm")
)
# Now we can feed the initial continuous and cateogrical layers to further
# "joint" layers after we concatenate them
net = (
ab.Concat(con_layer, cat_layer) >>
ab.Activation(tf.nn.selu) >>
ab.DenseVariational(output_dim=1)
)
# Split data into training and testing
Xt_con, Xs_con = np.split(X_con, [len(df_train)], axis=0)
Xt_cat, Xs_cat = np.split(X_cat, [len(df_train)], axis=0)
Yt, Ys = np.split(Y, [len(df_train)], axis=0)
# Graph place holders
X_con_ = tf.placeholder(tf.float32, [None, Xt_con.shape[1]])
X_cat_ = tf.placeholder(tf.int32, [None, Xt_cat.shape[1]])
Y_ = tf.placeholder(tf.float32, [None, 1])
# Feed dicts
train_dict = {X_con_: Xt_con, X_cat_: Xt_cat, Y_: Yt}
test_dict = {X_con_: Xs_con, X_cat_: Xs_cat, n_samples_: P_SAMPLES}
# Make model
N = len(Xt_con)
nn, kl = net(con=X_con_, cat=X_cat_)
likelihood = tf.distributions.Bernoulli(logits=nn)
prob = ab.sample_mean(likelihood.probs)
loss = ab.elbo(likelihood.log_prob(Y_), kl, N)
optimizer = tf.train.AdamOptimizer()
train = optimizer.minimize(loss)
init = tf.global_variables_initializer()
with tf.Session(config=CONFIG):
init.run()
# We're going to just use a feed_dict to feed in batches, which we
# generate here
batches = ab.batch(
train_dict,
batch_size=BSIZE,
n_iter=NITER)
for i, data in enumerate(batches):
train.run(feed_dict=data)
if i % 1000 == 0:
loss_val = loss.eval(feed_dict=data)
print("Iteration {}, loss = {}".format(i, loss_val))
# Predict
Ep = prob.eval(feed_dict=test_dict)
Ey = Ep > 0.5 # Max probability assignment
acc = accuracy_score(Ys.flatten(), Ey.flatten())
logloss = log_loss(Ys.flatten(), np.hstack((1 - Ep, Ep)))
print("Accuracy = {}, log loss = {}".format(acc, logloss))
# Network architecture
net = ab.stack(
ab.InputLayer(name='X', n_samples=l_samples), # LSAMPLES,BATCH_SIZE,28*28
ab.Conv2D(filters=32, kernel_size=(5, 5),
l2_reg=reg), # LSAMPLES, BATCH_SIZE, 28, 28, 32
ab.Activation(h=tf.nn.relu),
ab.MaxPool2D(pool_size=(2, 2),
strides=(2, 2)), # LSAMPLES, BATCH_SIZE, 14, 14, 32
ab.Conv2D(filters=64, kernel_size=(5, 5),
l2_reg=reg), # LSAMPLES, BATCH_SIZE, 14, 14, 64
ab.Activation(h=tf.nn.relu),
ab.MaxPool2D(pool_size=(2, 2),
strides=(2, 2)), # LSAMPLES, BATCH_SIZE, 7, 7, 64
ab.Flatten(), # LSAMPLES, BATCH_SIZE, 7*7*64
ab.Dense(output_dim=1024, l2_reg=reg), # LSAMPLES, BATCH_SIZE, 1024
ab.Activation(h=tf.nn.relu),
ab.DropOut(0.5),
ab.Dense(output_dim=10, l2_reg=reg), # LSAMPLES, BATCH_SIZE, 10
)
def main():
# Dataset
mnist_data = tf.contrib.learn.datasets.mnist.read_data_sets('./mnist_demo',
reshape=False)
N = mnist_data.train.images.shape[0]
X, Y = tf.data.Dataset.from_tensor_slices(
# Optimization
NITER = 20000 # Training iterations per fold
BSIZE = 10 # mini-batch size
CONFIG = tf.ConfigProto(device_count={'GPU': 0}) # Use GPU ?
LSAMPLES = 1 # We're only using 1 dropout "sample" for learning to be more
# like a MAP network
PSAMPLES = 50 # Number of samples for prediction
REG = 0.001 # weight regularizer
# Network structure
n_samples_ = tf.placeholder_with_default(LSAMPLES, [])
net = ab.stack(
ab.InputLayer(name='X', n_samples=n_samples_),
ab.DropOut(0.95, alpha=True),
ab.Dense(output_dim=128, l2_reg=REG, init_fn="autonorm"),
ab.Activation(h=tf.nn.selu),
ab.DropOut(0.9, alpha=True),
ab.Dense(output_dim=64, l2_reg=REG, init_fn="autonorm"),
ab.Activation(h=tf.nn.selu),
ab.DropOut(0.9, alpha=True),
ab.Dense(output_dim=32, l2_reg=REG, init_fn="autonorm"),
ab.Activation(h=tf.nn.selu),
ab.DropOut(0.9, alpha=True),
ab.Dense(output_dim=1, l2_reg=REG, init_fn="autonorm"),
)
def main():
"""Run the demo."""
data = load_breast_cancer()