 在 TensorFlow.org 上檢視
|
 在 Google Colab 中執行
|
 在 GitHub 上檢視原始碼
|
 下載筆記本
|
總覽
您可以使用 TFL Keras 層來建構具有單調性和其他形狀約束的 Keras 模型。此範例使用 TFL 層為 UCI 心臟資料集建立及訓練校正的格狀模型。
在校正的格狀模型中，每個特徵都透過 tfl.layers.PWLCalibration 或 tfl.layers.CategoricalCalibration 層轉換，結果會使用 tfl.layers.Lattice 進行非線性融合。
設定
安裝 TF Lattice 套件
pip install --pre -U tensorflow tf-keras tensorflow-lattice pydot graphviz
匯入必要的套件
import tensorflow as tf
import logging
import numpy as np
import pandas as pd
import sys
import tensorflow_lattice as tfl
from tensorflow import feature_column as fc
logging.disable(sys.maxsize)
# Use Keras 2.
version_fn = getattr(tf.keras, "version", None)
if version_fn and version_fn().startswith("3."):
import tf_keras as keras
else:
keras = tf.keras
下載 UCI Statlog (心臟) 資料集
# UCI Statlog (Heart) dataset.
csv_file = keras.utils.get_file(
'heart.csv', 'http://storage.googleapis.com/download.tensorflow.org/data/heart.csv')
training_data_df = pd.read_csv(csv_file).sample(
frac=1.0, random_state=41).reset_index(drop=True)
training_data_df.head()
Downloading data from http://storage.googleapis.com/download.tensorflow.org/data/heart.csv 13273/13273 [==============================] - 0s 0us/step
設定本指南中用於訓練的預設值
LEARNING_RATE = 0.1
BATCH_SIZE = 128
NUM_EPOCHS = 100
循序 Keras 模型
此範例建立循序 Keras 模型，且僅使用 TFL 層。
格狀層預期 input[i] 位於 [0, lattice_sizes[i] - 1.0] 範圍內，因此我們需要在校正層之前定義格狀大小，以便正確指定校正層的輸出範圍。
# Lattice layer expects input[i] to be within [0, lattice_sizes[i] - 1.0], so
lattice_sizes = [3, 2, 2, 2, 2, 2, 2]
我們使用 tfl.layers.ParallelCombination 層將必須平行執行的校正層分組在一起，以便能夠建立循序模型。
combined_calibrators = tfl.layers.ParallelCombination()
我們為每個特徵建立一個校正層，並將其新增至平行組合層。對於數值特徵，我們使用 tfl.layers.PWLCalibration，對於類別特徵，我們使用 tfl.layers.CategoricalCalibration。
# ############### age ###############
calibrator = tfl.layers.PWLCalibration(
# Every PWLCalibration layer must have keypoints of piecewise linear
# function specified. Easiest way to specify them is to uniformly cover
# entire input range by using numpy.linspace().
input_keypoints=np.linspace(
training_data_df['age'].min(), training_data_df['age'].max(), num=5),
# You need to ensure that input keypoints have same dtype as layer input.
# You can do it by setting dtype here or by providing keypoints in such
# format which will be converted to desired tf.dtype by default.
dtype=tf.float32,
# Output range must correspond to expected lattice input range.
output_min=0.0,
output_max=lattice_sizes[0] - 1.0,
)
combined_calibrators.append(calibrator)
# ############### sex ###############
# For boolean features simply specify CategoricalCalibration layer with 2
# buckets.
calibrator = tfl.layers.CategoricalCalibration(
num_buckets=2,
output_min=0.0,
output_max=lattice_sizes[1] - 1.0,
# Initializes all outputs to (output_min + output_max) / 2.0.
kernel_initializer='constant')
combined_calibrators.append(calibrator)
# ############### cp ###############
calibrator = tfl.layers.PWLCalibration(
# Here instead of specifying dtype of layer we convert keypoints into
# np.float32.
input_keypoints=np.linspace(1, 4, num=4, dtype=np.float32),
output_min=0.0,
output_max=lattice_sizes[2] - 1.0,
monotonicity='increasing',
# You can specify TFL regularizers as a tuple ('regularizer name', l1, l2).
kernel_regularizer=('hessian', 0.0, 1e-4))
combined_calibrators.append(calibrator)
# ############### trestbps ###############
calibrator = tfl.layers.PWLCalibration(
# Alternatively, you might want to use quantiles as keypoints instead of
# uniform keypoints
input_keypoints=np.quantile(training_data_df['trestbps'],
np.linspace(0.0, 1.0, num=5)),
dtype=tf.float32,
# Together with quantile keypoints you might want to initialize piecewise
# linear function to have 'equal_slopes' in order for output of layer
# after initialization to preserve original distribution.
kernel_initializer='equal_slopes',
output_min=0.0,
output_max=lattice_sizes[3] - 1.0,
# You might consider clamping extreme inputs of the calibrator to output
# bounds.
clamp_min=True,
clamp_max=True,
monotonicity='increasing')
combined_calibrators.append(calibrator)
# ############### chol ###############
calibrator = tfl.layers.PWLCalibration(
# Explicit input keypoint initialization.
input_keypoints=[126.0, 210.0, 247.0, 286.0, 564.0],
dtype=tf.float32,
output_min=0.0,
output_max=lattice_sizes[4] - 1.0,
# Monotonicity of calibrator can be decreasing. Note that corresponding
# lattice dimension must have INCREASING monotonicity regardless of
# monotonicity direction of calibrator.
monotonicity='decreasing',
# Convexity together with decreasing monotonicity result in diminishing
# return constraint.
convexity='convex',
# You can specify list of regularizers. You are not limited to TFL
# regularizrs. Feel free to use any :)
kernel_regularizer=[('laplacian', 0.0, 1e-4),
keras.regularizers.l1_l2(l1=0.001)])
combined_calibrators.append(calibrator)
# ############### fbs ###############
calibrator = tfl.layers.CategoricalCalibration(
num_buckets=2,
output_min=0.0,
output_max=lattice_sizes[5] - 1.0,
# For categorical calibration layer monotonicity is specified for pairs
# of indices of categories. Output for first category in pair will be
# smaller than output for second category.
#
# Don't forget to set monotonicity of corresponding dimension of Lattice
# layer to '1'.
monotonicities=[(0, 1)],
# This initializer is identical to default one('uniform'), but has fixed
# seed in order to simplify experimentation.
kernel_initializer=keras.initializers.RandomUniform(
minval=0.0, maxval=lattice_sizes[5] - 1.0, seed=1))
combined_calibrators.append(calibrator)
# ############### restecg ###############
calibrator = tfl.layers.CategoricalCalibration(
num_buckets=3,
output_min=0.0,
output_max=lattice_sizes[6] - 1.0,
# Categorical monotonicity can be partial order.
monotonicities=[(0, 1), (0, 2)],
# Categorical calibration layer supports standard Keras regularizers.
kernel_regularizer=keras.regularizers.l1_l2(l1=0.001),
kernel_initializer='constant')
combined_calibrators.append(calibrator)
然後，我們建立一個格狀層，以非線性方式融合校正器的輸出。
請注意，我們需要指定格狀的單調性在所需的維度上是遞增的。與校正中單調性方向的組合將產生正確的端對端單調性方向。這包括 CategoricalCalibration 層的部分單調性。
lattice = tfl.layers.Lattice(
lattice_sizes=lattice_sizes,
monotonicities=[
'increasing', 'none', 'increasing', 'increasing', 'increasing',
'increasing', 'increasing'
],
output_min=0.0,
output_max=1.0)
然後，我們可以使用組合的校正器和格狀層來建立循序模型。
model = keras.models.Sequential()
model.add(combined_calibrators)
model.add(lattice)
2024-03-23 11:18:06.857910: E external/local_xla/xla/stream_executor/cuda/cuda_driver.cc:282] failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected
訓練方式與任何其他 Keras 模型相同。
features = training_data_df[[
'age', 'sex', 'cp', 'trestbps', 'chol', 'fbs', 'restecg'
]].values.astype(np.float32)
target = training_data_df[['target']].values.astype(np.float32)
model.compile(
loss=keras.losses.mean_squared_error,
optimizer=keras.optimizers.Adagrad(learning_rate=LEARNING_RATE))
model.fit(
features,
target,
batch_size=BATCH_SIZE,
epochs=NUM_EPOCHS,
validation_split=0.2,
shuffle=False,
verbose=0)
model.evaluate(features, target)
10/10 [==============================] - 0s 2ms/step - loss: 0.1551 0.15509344637393951
函數式 Keras 模型
此範例使用函數式 API 進行 Keras 模型建構。
如前一節所述，格狀層預期 input[i] 位於 [0, lattice_sizes[i] - 1.0] 範圍內，因此我們需要在校正層之前定義格狀大小，以便正確指定校正層的輸出範圍。
# We are going to have 2-d embedding as one of lattice inputs.
lattice_sizes = [3, 2, 2, 3, 3, 2, 2]
對於每個特徵，我們需要建立一個輸入層，然後建立一個校正層。對於數值特徵，我們使用 tfl.layers.PWLCalibration，對於類別特徵，我們使用 tfl.layers.CategoricalCalibration。
model_inputs = []
lattice_inputs = []
# ############### age ###############
age_input = keras.layers.Input(shape=[1], name='age')
model_inputs.append(age_input)
age_calibrator = tfl.layers.PWLCalibration(
# Every PWLCalibration layer must have keypoints of piecewise linear
# function specified. Easiest way to specify them is to uniformly cover
# entire input range by using numpy.linspace().
input_keypoints=np.linspace(
training_data_df['age'].min(), training_data_df['age'].max(), num=5),
# You need to ensure that input keypoints have same dtype as layer input.
# You can do it by setting dtype here or by providing keypoints in such
# format which will be converted to desired tf.dtype by default.
dtype=tf.float32,
# Output range must correspond to expected lattice input range.
output_min=0.0,
output_max=lattice_sizes[0] - 1.0,
monotonicity='increasing',
name='age_calib',
)(
age_input)
lattice_inputs.append(age_calibrator)
# ############### sex ###############
# For boolean features simply specify CategoricalCalibration layer with 2
# buckets.
sex_input = keras.layers.Input(shape=[1], name='sex')
model_inputs.append(sex_input)
sex_calibrator = tfl.layers.CategoricalCalibration(
num_buckets=2,
output_min=0.0,
output_max=lattice_sizes[1] - 1.0,
# Initializes all outputs to (output_min + output_max) / 2.0.
kernel_initializer='constant',
name='sex_calib',
)(
sex_input)
lattice_inputs.append(sex_calibrator)
# ############### cp ###############
cp_input = keras.layers.Input(shape=[1], name='cp')
model_inputs.append(cp_input)
cp_calibrator = tfl.layers.PWLCalibration(
# Here instead of specifying dtype of layer we convert keypoints into
# np.float32.
input_keypoints=np.linspace(1, 4, num=4, dtype=np.float32),
output_min=0.0,
output_max=lattice_sizes[2] - 1.0,
monotonicity='increasing',
# You can specify TFL regularizers as tuple ('regularizer name', l1, l2).
kernel_regularizer=('hessian', 0.0, 1e-4),
name='cp_calib',
)(
cp_input)
lattice_inputs.append(cp_calibrator)
# ############### trestbps ###############
trestbps_input = keras.layers.Input(shape=[1], name='trestbps')
model_inputs.append(trestbps_input)
trestbps_calibrator = tfl.layers.PWLCalibration(
# Alternatively, you might want to use quantiles as keypoints instead of
# uniform keypoints
input_keypoints=np.quantile(training_data_df['trestbps'],
np.linspace(0.0, 1.0, num=5)),
dtype=tf.float32,
# Together with quantile keypoints you might want to initialize piecewise
# linear function to have 'equal_slopes' in order for output of layer
# after initialization to preserve original distribution.
kernel_initializer='equal_slopes',
output_min=0.0,
output_max=lattice_sizes[3] - 1.0,
# You might consider clamping extreme inputs of the calibrator to output
# bounds.
clamp_min=True,
clamp_max=True,
monotonicity='increasing',
name='trestbps_calib',
)(
trestbps_input)
lattice_inputs.append(trestbps_calibrator)
# ############### chol ###############
chol_input = keras.layers.Input(shape=[1], name='chol')
model_inputs.append(chol_input)
chol_calibrator = tfl.layers.PWLCalibration(
# Explicit input keypoint initialization.
input_keypoints=[126.0, 210.0, 247.0, 286.0, 564.0],
output_min=0.0,
output_max=lattice_sizes[4] - 1.0,
# Monotonicity of calibrator can be decreasing. Note that corresponding
# lattice dimension must have INCREASING monotonicity regardless of
# monotonicity direction of calibrator.
monotonicity='decreasing',
# Convexity together with decreasing monotonicity result in diminishing
# return constraint.
convexity='convex',
# You can specify list of regularizers. You are not limited to TFL
# regularizrs. Feel free to use any :)
kernel_regularizer=[('laplacian', 0.0, 1e-4),
keras.regularizers.l1_l2(l1=0.001)],
name='chol_calib',
)(
chol_input)
lattice_inputs.append(chol_calibrator)
# ############### fbs ###############
fbs_input = keras.layers.Input(shape=[1], name='fbs')
model_inputs.append(fbs_input)
fbs_calibrator = tfl.layers.CategoricalCalibration(
num_buckets=2,
output_min=0.0,
output_max=lattice_sizes[5] - 1.0,
# For categorical calibration layer monotonicity is specified for pairs
# of indices of categories. Output for first category in pair will be
# smaller than output for second category.
#
# Don't forget to set monotonicity of corresponding dimension of Lattice
# layer to '1'.
monotonicities=[(0, 1)],
# This initializer is identical to default one ('uniform'), but has fixed
# seed in order to simplify experimentation.
kernel_initializer=keras.initializers.RandomUniform(
minval=0.0, maxval=lattice_sizes[5] - 1.0, seed=1),
name='fbs_calib',
)(
fbs_input)
lattice_inputs.append(fbs_calibrator)
# ############### restecg ###############
restecg_input = keras.layers.Input(shape=[1], name='restecg')
model_inputs.append(restecg_input)
restecg_calibrator = tfl.layers.CategoricalCalibration(
num_buckets=3,
output_min=0.0,
output_max=lattice_sizes[6] - 1.0,
# Categorical monotonicity can be partial order.
monotonicities=[(0, 1), (0, 2)],
# Categorical calibration layer supports standard Keras regularizers.
kernel_regularizer=keras.regularizers.l1_l2(l1=0.001),
kernel_initializer='constant',
name='restecg_calib',
)(
restecg_input)
lattice_inputs.append(restecg_calibrator)
然後，我們建立一個格狀層，以非線性方式融合校正器的輸出。
請注意，我們需要指定格狀的單調性在所需的維度上是遞增的。與校正中單調性方向的組合將產生正確的端對端單調性方向。這包括 tfl.layers.CategoricalCalibration 層的部分單調性。
lattice = tfl.layers.Lattice(
lattice_sizes=lattice_sizes,
monotonicities=[
'increasing', 'none', 'increasing', 'increasing', 'increasing',
'increasing', 'increasing'
],
output_min=0.0,
output_max=1.0,
name='lattice',
)(
lattice_inputs)
為了增加模型的彈性，我們新增一個輸出校正層。
model_output = tfl.layers.PWLCalibration(
input_keypoints=np.linspace(0.0, 1.0, 5),
name='output_calib',
)(
lattice)
我們現在可以使用輸入和輸出建立模型。
model = keras.models.Model(
inputs=model_inputs,
outputs=model_output)
keras.utils.plot_model(model, rankdir='LR')
訓練方式與任何其他 Keras 模型相同。請注意，在我們的設定中，輸入特徵會以個別張量的形式傳遞。
feature_names = ['age', 'sex', 'cp', 'trestbps', 'chol', 'fbs', 'restecg']
features = np.split(
training_data_df[feature_names].values.astype(np.float32),
indices_or_sections=len(feature_names),
axis=1)
target = training_data_df[['target']].values.astype(np.float32)
model.compile(
loss=keras.losses.mean_squared_error,
optimizer=keras.optimizers.Adagrad(LEARNING_RATE))
model.fit(
features,
target,
batch_size=BATCH_SIZE,
epochs=NUM_EPOCHS,
validation_split=0.2,
shuffle=False,
verbose=0)
model.evaluate(features, target)
10/10 [==============================] - 0s 2ms/step - loss: 0.1580 0.15800504386425018