Lihat di TensorFlow.org | Jalankan di Google Colab | Lihat sumber di GitHub | Unduh buku catatan |
Ringkasan
Premade Model yang cepat dan mudah cara untuk membangun TFL tf.keras.model
contoh untuk kasus penggunaan yang khas. Panduan ini menguraikan langkah-langkah yang diperlukan untuk membuat Model TFL Premade dan melatih/mengujinya.
Mempersiapkan
Menginstal paket TF Lattice:
pip install tensorflow-lattice pydot
Mengimpor paket yang dibutuhkan:
import tensorflow as tf
import copy
import logging
import numpy as np
import pandas as pd
import sys
import tensorflow_lattice as tfl
logging.disable(sys.maxsize)
Menyetel nilai default yang digunakan untuk pelatihan dalam panduan ini:
LEARNING_RATE = 0.01
BATCH_SIZE = 128
NUM_EPOCHS = 500
PREFITTING_NUM_EPOCHS = 10
Mengunduh kumpulan data UCI Statlog (Hati):
heart_csv_file = tf.keras.utils.get_file(
'heart.csv',
'http://storage.googleapis.com/download.tensorflow.org/data/heart.csv')
heart_df = pd.read_csv(heart_csv_file)
thal_vocab_list = ['normal', 'fixed', 'reversible']
heart_df['thal'] = heart_df['thal'].map(
{v: i for i, v in enumerate(thal_vocab_list)})
heart_df = heart_df.astype(float)
heart_train_size = int(len(heart_df) * 0.8)
heart_train_dict = dict(heart_df[:heart_train_size])
heart_test_dict = dict(heart_df[heart_train_size:])
# This ordering of input features should match the feature configs. If no
# feature config relies explicitly on the data (i.e. all are 'quantiles'),
# then you can construct the feature_names list by simply iterating over each
# feature config and extracting it's name.
feature_names = [
'age', 'sex', 'cp', 'chol', 'fbs', 'trestbps', 'thalach', 'restecg',
'exang', 'oldpeak', 'slope', 'ca', 'thal'
]
# Since we have some features that manually construct their input keypoints,
# we need an index mapping of the feature names.
feature_name_indices = {name: index for index, name in enumerate(feature_names)}
label_name = 'target'
heart_train_xs = [
heart_train_dict[feature_name] for feature_name in feature_names
]
heart_test_xs = [heart_test_dict[feature_name] for feature_name in feature_names]
heart_train_ys = heart_train_dict[label_name]
heart_test_ys = heart_test_dict[label_name]
Downloading data from http://storage.googleapis.com/download.tensorflow.org/data/heart.csv 16384/13273 [=====================================] - 0s 0us/step 24576/13273 [=======================================================] - 0s 0us/step
Konfigurasi Fitur
Fitur kalibrasi dan konfigurasi per-fitur diatur menggunakan tfl.configs.FeatureConfig . Konfigurasi fitur termasuk kendala monotonisitas, regularisasi per-fitur (lihat tfl.configs.RegularizerConfig ), dan ukuran kisi untuk model kisi.
Perhatikan bahwa kita harus sepenuhnya menentukan konfigurasi fitur untuk fitur apa pun yang kita ingin model kita kenali. Jika tidak, model tidak akan mengetahui bahwa fitur seperti itu ada.
Mendefinisikan Konfigurasi Fitur Kami
Sekarang setelah kami dapat menghitung kuantil kami, kami mendefinisikan konfigurasi fitur untuk setiap fitur yang kami ingin model kami ambil sebagai input.
# Features:
# - age
# - sex
# - cp chest pain type (4 values)
# - trestbps resting blood pressure
# - chol serum cholestoral in mg/dl
# - fbs fasting blood sugar > 120 mg/dl
# - restecg resting electrocardiographic results (values 0,1,2)
# - thalach maximum heart rate achieved
# - exang exercise induced angina
# - oldpeak ST depression induced by exercise relative to rest
# - slope the slope of the peak exercise ST segment
# - ca number of major vessels (0-3) colored by flourosopy
# - thal normal; fixed defect; reversable defect
#
# Feature configs are used to specify how each feature is calibrated and used.
heart_feature_configs = [
tfl.configs.FeatureConfig(
name='age',
lattice_size=3,
monotonicity='increasing',
# We must set the keypoints manually.
pwl_calibration_num_keypoints=5,
pwl_calibration_input_keypoints='quantiles',
pwl_calibration_clip_max=100,
# Per feature regularization.
regularizer_configs=[
tfl.configs.RegularizerConfig(name='calib_wrinkle', l2=0.1),
],
),
tfl.configs.FeatureConfig(
name='sex',
num_buckets=2,
),
tfl.configs.FeatureConfig(
name='cp',
monotonicity='increasing',
# Keypoints that are uniformly spaced.
pwl_calibration_num_keypoints=4,
pwl_calibration_input_keypoints=np.linspace(
np.min(heart_train_xs[feature_name_indices['cp']]),
np.max(heart_train_xs[feature_name_indices['cp']]),
num=4),
),
tfl.configs.FeatureConfig(
name='chol',
monotonicity='increasing',
# Explicit input keypoints initialization.
pwl_calibration_input_keypoints=[126.0, 210.0, 247.0, 286.0, 564.0],
# Calibration can be forced to span the full output range by clamping.
pwl_calibration_clamp_min=True,
pwl_calibration_clamp_max=True,
# Per feature regularization.
regularizer_configs=[
tfl.configs.RegularizerConfig(name='calib_hessian', l2=1e-4),
],
),
tfl.configs.FeatureConfig(
name='fbs',
# Partial monotonicity: output(0) <= output(1)
monotonicity=[(0, 1)],
num_buckets=2,
),
tfl.configs.FeatureConfig(
name='trestbps',
monotonicity='decreasing',
pwl_calibration_num_keypoints=5,
pwl_calibration_input_keypoints='quantiles',
),
tfl.configs.FeatureConfig(
name='thalach',
monotonicity='decreasing',
pwl_calibration_num_keypoints=5,
pwl_calibration_input_keypoints='quantiles',
),
tfl.configs.FeatureConfig(
name='restecg',
# Partial monotonicity: output(0) <= output(1), output(0) <= output(2)
monotonicity=[(0, 1), (0, 2)],
num_buckets=3,
),
tfl.configs.FeatureConfig(
name='exang',
# Partial monotonicity: output(0) <= output(1)
monotonicity=[(0, 1)],
num_buckets=2,
),
tfl.configs.FeatureConfig(
name='oldpeak',
monotonicity='increasing',
pwl_calibration_num_keypoints=5,
pwl_calibration_input_keypoints='quantiles',
),
tfl.configs.FeatureConfig(
name='slope',
# Partial monotonicity: output(0) <= output(1), output(1) <= output(2)
monotonicity=[(0, 1), (1, 2)],
num_buckets=3,
),
tfl.configs.FeatureConfig(
name='ca',
monotonicity='increasing',
pwl_calibration_num_keypoints=4,
pwl_calibration_input_keypoints='quantiles',
),
tfl.configs.FeatureConfig(
name='thal',
# Partial monotonicity:
# output(normal) <= output(fixed)
# output(normal) <= output(reversible)
monotonicity=[('normal', 'fixed'), ('normal', 'reversible')],
num_buckets=3,
# We must specify the vocabulary list in order to later set the
# monotonicities since we used names and not indices.
vocabulary_list=thal_vocab_list,
),
]
Tetapkan Monotonisitas dan Poin Utama
Selanjutnya kita perlu memastikan untuk mengatur monotonisitas dengan benar untuk fitur di mana kita menggunakan kosakata khusus (seperti 'thal' di atas).
tfl.premade_lib.set_categorical_monotonicities(heart_feature_configs)
Akhirnya kami dapat menyelesaikan konfigurasi fitur kami dengan menghitung dan mengatur titik kunci.
feature_keypoints = tfl.premade_lib.compute_feature_keypoints(
feature_configs=heart_feature_configs, features=heart_train_dict)
tfl.premade_lib.set_feature_keypoints(
feature_configs=heart_feature_configs,
feature_keypoints=feature_keypoints,
add_missing_feature_configs=False)
Model Linier Terkalibrasi
Untuk membangun sebuah model premade TFL, pertama membangun sebuah konfigurasi Model dari tfl.configs . Sebuah model linear dikalibrasi dibangun menggunakan tfl.configs.CalibratedLinearConfig . Ini menerapkan kalibrasi piecewise-linear dan kategoris pada fitur input, diikuti dengan kombinasi linier dan kalibrasi output piecewise-linear opsional. Saat menggunakan kalibrasi keluaran atau ketika batas keluaran ditentukan, lapisan linier akan menerapkan rata-rata tertimbang pada masukan yang dikalibrasi.
Contoh ini membuat model linier terkalibrasi pada 5 fitur pertama.
# Model config defines the model structure for the premade model.
linear_model_config = tfl.configs.CalibratedLinearConfig(
feature_configs=heart_feature_configs[:5],
use_bias=True,
output_calibration=True,
output_calibration_num_keypoints=10,
# We initialize the output to [-2.0, 2.0] since we'll be using logits.
output_initialization=np.linspace(-2.0, 2.0, num=10),
regularizer_configs=[
# Regularizer for the output calibrator.
tfl.configs.RegularizerConfig(name='output_calib_hessian', l2=1e-4),
])
# A CalibratedLinear premade model constructed from the given model config.
linear_model = tfl.premade.CalibratedLinear(linear_model_config)
# Let's plot our model.
tf.keras.utils.plot_model(linear_model, show_layer_names=False, rankdir='LR')
2022-01-14 12:36:31.295751: E tensorflow/stream_executor/cuda/cuda_driver.cc:271] failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected
Sekarang, karena dengan lainnya tf.keras.Model , kita mengkompilasi dan sesuai dengan model data kami.
linear_model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
metrics=[tf.keras.metrics.AUC(from_logits=True)],
optimizer=tf.keras.optimizers.Adam(LEARNING_RATE))
linear_model.fit(
heart_train_xs[:5],
heart_train_ys,
epochs=NUM_EPOCHS,
batch_size=BATCH_SIZE,
verbose=False)
<keras.callbacks.History at 0x7fe4385f0290>
Setelah melatih model kami, kami dapat mengevaluasinya di set pengujian kami.
print('Test Set Evaluation...')
print(linear_model.evaluate(heart_test_xs[:5], heart_test_ys))
Test Set Evaluation... 2/2 [==============================] - 0s 3ms/step - loss: 0.4728 - auc: 0.8252 [0.47278329730033875, 0.8251879215240479]
Model Kisi Terkalibrasi
Sebuah model kisi dikalibrasi dibangun menggunakan tfl.configs.CalibratedLatticeConfig . Model kisi yang dikalibrasi menerapkan kalibrasi linier sepotong-sepotong dan kategoris pada fitur input, diikuti oleh model kisi dan kalibrasi linier sepotong-sepotong keluaran opsional.
Contoh ini membuat model kisi yang dikalibrasi pada 5 fitur pertama.
# This is a calibrated lattice model: inputs are calibrated, then combined
# non-linearly using a lattice layer.
lattice_model_config = tfl.configs.CalibratedLatticeConfig(
feature_configs=heart_feature_configs[:5],
# We initialize the output to [-2.0, 2.0] since we'll be using logits.
output_initialization=[-2.0, 2.0],
regularizer_configs=[
# Torsion regularizer applied to the lattice to make it more linear.
tfl.configs.RegularizerConfig(name='torsion', l2=1e-2),
# Globally defined calibration regularizer is applied to all features.
tfl.configs.RegularizerConfig(name='calib_hessian', l2=1e-2),
])
# A CalibratedLattice premade model constructed from the given model config.
lattice_model = tfl.premade.CalibratedLattice(lattice_model_config)
# Let's plot our model.
tf.keras.utils.plot_model(lattice_model, show_layer_names=False, rankdir='LR')
Seperti sebelumnya, kami menyusun, menyesuaikan, dan mengevaluasi model kami.
lattice_model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
metrics=[tf.keras.metrics.AUC(from_logits=True)],
optimizer=tf.keras.optimizers.Adam(LEARNING_RATE))
lattice_model.fit(
heart_train_xs[:5],
heart_train_ys,
epochs=NUM_EPOCHS,
batch_size=BATCH_SIZE,
verbose=False)
print('Test Set Evaluation...')
print(lattice_model.evaluate(heart_test_xs[:5], heart_test_ys))
Test Set Evaluation... 2/2 [==============================] - 1s 3ms/step - loss: 0.4709 - auc_1: 0.8302 [0.4709009826183319, 0.8302004933357239]
Model Ensemble Kisi Terkalibrasi
Ketika jumlah fitur besar, Anda dapat menggunakan model ensemble, yang membuat beberapa kisi yang lebih kecil untuk subset fitur dan rata-rata outputnya alih-alih membuat hanya satu kisi besar. Model kisi Ensemble dibangun menggunakan tfl.configs.CalibratedLatticeEnsembleConfig . Model ansambel kisi yang dikalibrasi menerapkan kalibrasi linier-sepotong dan kategoris pada fitur input, diikuti oleh ansambel model kisi dan kalibrasi linier-sepotong output opsional.
Inisialisasi Ensemble Kisi Eksplisit
Jika Anda sudah mengetahui subkumpulan fitur mana yang ingin Anda masukkan ke dalam kisi, Anda dapat menyetel kisi secara eksplisit menggunakan nama fitur. Contoh ini membuat model ensembel kisi terkalibrasi dengan 5 kisi dan 3 fitur per kisi.
# This is a calibrated lattice ensemble model: inputs are calibrated, then
# combined non-linearly and averaged using multiple lattice layers.
explicit_ensemble_model_config = tfl.configs.CalibratedLatticeEnsembleConfig(
feature_configs=heart_feature_configs,
lattices=[['trestbps', 'chol', 'ca'], ['fbs', 'restecg', 'thal'],
['fbs', 'cp', 'oldpeak'], ['exang', 'slope', 'thalach'],
['restecg', 'age', 'sex']],
num_lattices=5,
lattice_rank=3,
# We initialize the output to [-2.0, 2.0] since we'll be using logits.
output_initialization=[-2.0, 2.0])
# A CalibratedLatticeEnsemble premade model constructed from the given
# model config.
explicit_ensemble_model = tfl.premade.CalibratedLatticeEnsemble(
explicit_ensemble_model_config)
# Let's plot our model.
tf.keras.utils.plot_model(
explicit_ensemble_model, show_layer_names=False, rankdir='LR')
Seperti sebelumnya, kami menyusun, menyesuaikan, dan mengevaluasi model kami.
explicit_ensemble_model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
metrics=[tf.keras.metrics.AUC(from_logits=True)],
optimizer=tf.keras.optimizers.Adam(LEARNING_RATE))
explicit_ensemble_model.fit(
heart_train_xs,
heart_train_ys,
epochs=NUM_EPOCHS,
batch_size=BATCH_SIZE,
verbose=False)
print('Test Set Evaluation...')
print(explicit_ensemble_model.evaluate(heart_test_xs, heart_test_ys))
Test Set Evaluation... 2/2 [==============================] - 1s 4ms/step - loss: 0.3768 - auc_2: 0.8954 [0.3768467903137207, 0.895363450050354]
Ensemble Kisi Acak
Jika Anda tidak yakin subkumpulan fitur mana yang akan dimasukkan ke dalam kisi Anda, opsi lainnya adalah menggunakan subkumpulan fitur acak untuk setiap kisi. Contoh ini membuat model ensembel kisi terkalibrasi dengan 5 kisi dan 3 fitur per kisi.
# This is a calibrated lattice ensemble model: inputs are calibrated, then
# combined non-linearly and averaged using multiple lattice layers.
random_ensemble_model_config = tfl.configs.CalibratedLatticeEnsembleConfig(
feature_configs=heart_feature_configs,
lattices='random',
num_lattices=5,
lattice_rank=3,
# We initialize the output to [-2.0, 2.0] since we'll be using logits.
output_initialization=[-2.0, 2.0],
random_seed=42)
# Now we must set the random lattice structure and construct the model.
tfl.premade_lib.set_random_lattice_ensemble(random_ensemble_model_config)
# A CalibratedLatticeEnsemble premade model constructed from the given
# model config.
random_ensemble_model = tfl.premade.CalibratedLatticeEnsemble(
random_ensemble_model_config)
# Let's plot our model.
tf.keras.utils.plot_model(
random_ensemble_model, show_layer_names=False, rankdir='LR')
Seperti sebelumnya, kami menyusun, menyesuaikan, dan mengevaluasi model kami.
random_ensemble_model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
metrics=[tf.keras.metrics.AUC(from_logits=True)],
optimizer=tf.keras.optimizers.Adam(LEARNING_RATE))
random_ensemble_model.fit(
heart_train_xs,
heart_train_ys,
epochs=NUM_EPOCHS,
batch_size=BATCH_SIZE,
verbose=False)
print('Test Set Evaluation...')
print(random_ensemble_model.evaluate(heart_test_xs, heart_test_ys))
Test Set Evaluation... 2/2 [==============================] - 1s 4ms/step - loss: 0.3739 - auc_3: 0.8997 [0.3739270567893982, 0.8997493982315063]
Ensemble Kisi Acak Lapisan RTL
Bila menggunakan kisi acak ansambel, Anda dapat menentukan bahwa model menggunakan satu tfl.layers.RTL
lapisan. Kami mencatat bahwa tfl.layers.RTL
hanya mendukung kendala monotonisitas dan harus memiliki ukuran kisi yang sama untuk semua fitur dan tidak ada regularisasi per-fitur. Perhatikan bahwa menggunakan tfl.layers.RTL
lapisan memungkinkan Anda skala untuk ansambel jauh lebih besar daripada menggunakan terpisah tfl.layers.Lattice
contoh.
Contoh ini membuat model ensembel kisi terkalibrasi dengan 5 kisi dan 3 fitur per kisi.
# Make sure our feature configs have the same lattice size, no per-feature
# regularization, and only monotonicity constraints.
rtl_layer_feature_configs = copy.deepcopy(heart_feature_configs)
for feature_config in rtl_layer_feature_configs:
feature_config.lattice_size = 2
feature_config.unimodality = 'none'
feature_config.reflects_trust_in = None
feature_config.dominates = None
feature_config.regularizer_configs = None
# This is a calibrated lattice ensemble model: inputs are calibrated, then
# combined non-linearly and averaged using multiple lattice layers.
rtl_layer_ensemble_model_config = tfl.configs.CalibratedLatticeEnsembleConfig(
feature_configs=rtl_layer_feature_configs,
lattices='rtl_layer',
num_lattices=5,
lattice_rank=3,
# We initialize the output to [-2.0, 2.0] since we'll be using logits.
output_initialization=[-2.0, 2.0],
random_seed=42)
# A CalibratedLatticeEnsemble premade model constructed from the given
# model config. Note that we do not have to specify the lattices by calling
# a helper function (like before with random) because the RTL Layer will take
# care of that for us.
rtl_layer_ensemble_model = tfl.premade.CalibratedLatticeEnsemble(
rtl_layer_ensemble_model_config)
# Let's plot our model.
tf.keras.utils.plot_model(
rtl_layer_ensemble_model, show_layer_names=False, rankdir='LR')
Seperti sebelumnya, kami menyusun, menyesuaikan, dan mengevaluasi model kami.
rtl_layer_ensemble_model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
metrics=[tf.keras.metrics.AUC(from_logits=True)],
optimizer=tf.keras.optimizers.Adam(LEARNING_RATE))
rtl_layer_ensemble_model.fit(
heart_train_xs,
heart_train_ys,
epochs=NUM_EPOCHS,
batch_size=BATCH_SIZE,
verbose=False)
print('Test Set Evaluation...')
print(rtl_layer_ensemble_model.evaluate(heart_test_xs, heart_test_ys))
Test Set Evaluation... 2/2 [==============================] - 0s 3ms/step - loss: 0.3614 - auc_4: 0.9079 [0.36142951250076294, 0.9078947305679321]
Ansambel Kisi Kristal
Premade juga menyediakan algoritma pengaturan fitur heuristik, yang disebut Kristal . Untuk menggunakan algoritma Crystals, pertama-tama kita melatih model prefitting yang memperkirakan interaksi fitur berpasangan. Kami kemudian mengatur ansambel akhir sedemikian rupa sehingga fitur dengan lebih banyak interaksi non-linier berada dalam kisi yang sama.
Perpustakaan Premade menawarkan fungsi pembantu untuk membangun konfigurasi model prefitting dan mengekstraksi struktur kristal. Perhatikan bahwa model prefitting tidak perlu dilatih sepenuhnya, jadi beberapa epoch sudah cukup.
Contoh ini membuat model ensembel kisi terkalibrasi dengan 5 kisi dan 3 fitur per kisi.
# This is a calibrated lattice ensemble model: inputs are calibrated, then
# combines non-linearly and averaged using multiple lattice layers.
crystals_ensemble_model_config = tfl.configs.CalibratedLatticeEnsembleConfig(
feature_configs=heart_feature_configs,
lattices='crystals',
num_lattices=5,
lattice_rank=3,
# We initialize the output to [-2.0, 2.0] since we'll be using logits.
output_initialization=[-2.0, 2.0],
random_seed=42)
# Now that we have our model config, we can construct a prefitting model config.
prefitting_model_config = tfl.premade_lib.construct_prefitting_model_config(
crystals_ensemble_model_config)
# A CalibratedLatticeEnsemble premade model constructed from the given
# prefitting model config.
prefitting_model = tfl.premade.CalibratedLatticeEnsemble(
prefitting_model_config)
# We can compile and train our prefitting model as we like.
prefitting_model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
optimizer=tf.keras.optimizers.Adam(LEARNING_RATE))
prefitting_model.fit(
heart_train_xs,
heart_train_ys,
epochs=PREFITTING_NUM_EPOCHS,
batch_size=BATCH_SIZE,
verbose=False)
# Now that we have our trained prefitting model, we can extract the crystals.
tfl.premade_lib.set_crystals_lattice_ensemble(crystals_ensemble_model_config,
prefitting_model_config,
prefitting_model)
# A CalibratedLatticeEnsemble premade model constructed from the given
# model config.
crystals_ensemble_model = tfl.premade.CalibratedLatticeEnsemble(
crystals_ensemble_model_config)
# Let's plot our model.
tf.keras.utils.plot_model(
crystals_ensemble_model, show_layer_names=False, rankdir='LR')
Seperti sebelumnya, kami menyusun, menyesuaikan, dan mengevaluasi model kami.
crystals_ensemble_model.compile(
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
metrics=[tf.keras.metrics.AUC(from_logits=True)],
optimizer=tf.keras.optimizers.Adam(LEARNING_RATE))
crystals_ensemble_model.fit(
heart_train_xs,
heart_train_ys,
epochs=NUM_EPOCHS,
batch_size=BATCH_SIZE,
verbose=False)
print('Test Set Evaluation...')
print(crystals_ensemble_model.evaluate(heart_test_xs, heart_test_ys))
Test Set Evaluation... 2/2 [==============================] - 1s 3ms/step - loss: 0.3404 - auc_5: 0.9179 [0.34039050340652466, 0.9179198145866394]