AttributeError: 'Sequential' object has no attribute 'predict_proba'

Question

predict_proba returns the error in the neural network

i saw the example on this link https://machinelearningmastery.com/how-to-make-classification-and-regression-predictions-for-deep-learning-models-in-keras/

https://faroit.com/keras-docs/1.0.0/models/sequential/#the-sequential-model-api

I am using Tensorflow Version: 2.6.0

Code:

#creating the object (Initializing the ANN)
import tensorflow as tf
from tensorflow import keras
LAYERS = [
         tf.keras.layers.Dense(50, activation="relu", input_shape=X_train.shape[1:]),
         tf.keras.layers.LeakyReLU(),
         tf.keras.layers.Dense(25, activation="relu"),
         tf.keras.layers.Dense(10, activation="relu"),
         tf.keras.layers.Dense(5, activation="relu"),
         tf.keras.layers.Flatten(),
         tf.keras.layers.Dense(1, activation='sigmoid')
]

LOSS = "binary_crossentropy"
OPTIMIZER = tf.keras.optimizers.Adam(learning_rate=1e-3)

model_cEXT = tf.keras.models.Sequential(LAYERS)
model_cEXT.compile(loss=LOSS , optimizer=OPTIMIZER, metrics=['accuracy'])

EPOCHS = 100

checkpoint_cb = tf.keras.callbacks.ModelCheckpoint("model_cEXT.h5", save_best_only=True)
early_stopping_cb = tf.keras.callbacks.EarlyStopping(patience=10, restore_best_weights=True)
tensorboard_cb = tf.keras.callbacks.TensorBoard(log_dir="logs")
CALLBACKS = [checkpoint_cb, early_stopping_cb, tensorboard_cb]

model_cEXT.fit(X_train, y_train['cEXT'], epochs = EPOCHS, validation_data=(X_test, y_test['cEXT']), callbacks = CALLBACKS)

model_cEXT.predict_proba(X_test)

Error:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-72-8f06353cf345> in <module>()
----> 1 model_cEXT.predict_proba(X_test)

AttributeError: 'Sequential' object has no attribute 'predict_proba'

Edit: i need sklearn's like predict_proba output it is needed for visualization

skplt.metrics.plot_precision_recall_curve(y_test['cEXT'].values, y_prob)
plt.title('Precision-Recall Curve - cEXT')
plt.show()

What is your question? The error is clear. Tensorflow is not scikit-learn. — xdurch0
– xdurch0, Commented Aug 29, 2021 at 9:02
wants to print probability like scikit-learn's predict_proba — Juned Ansari
– Juned Ansari, Commented Aug 29, 2021 at 9:04
but i saw the example on deep learning too machinelearningmastery.com/… — Juned Ansari
– Juned Ansari, Commented Aug 29, 2021 at 9:05
@JunedAnsari predict_proba is deprecated. Instead, simply use predict to get the probabilities. See my answer here for more details and let me know if it helps. stackoverflow.com/a/67467084/9215780 — Innat
– Innat, Commented Aug 29, 2021 at 14:15
Huh, my apologies. I wasn't aware old keras versions had this method. — xdurch0
– xdurch0, Commented Aug 29, 2021 at 20:03

m4n0 · Accepted Answer · 2021-09-28 06:33:24Z

7

Use this code instead

predict_prob=model.predict([testa,testb])

predict_classes=np.argmax(predict_prob,axis=1)

edited Sep 28, 2021 at 6:33

m4n0

32.6k28 gold badges81 silver badges98 bronze badges

answered Sep 28, 2021 at 5:49

Supriya

711 bronze badge

Sign up to request clarification or add additional context in comments.

Comments

Juned Ansari · Accepted Answer · 2021-08-30 02:52:09Z

1

New Version might not have predict_proba method so i have creadted my own using .predict method

def predict_prob(number):
  return [number[0],1-number[0]]

y_prob = np.array(list(map(predict_prob, model_cEXT.predict(X_test))))
y_prob

answered Aug 30, 2021 at 2:52

Juned Ansari

5,3137 gold badges69 silver badges92 bronze badges

2 Comments

Innat Over a year ago

Note, It's NOT the solution. The mode. predict already gives the probabilities. The new version is more precise. In your above code snippet, it doesn't make sense to compute twice the probabilities of the same class.

Juned Ansari Over a year ago

@M.Innat, i know it returns probability, but if you compare it with sklearns predict_prob you will see the difference. sklearns prdict_prob will return two output like true class probability and the false class probability it was needed for the visualization skplt.metrics.plot_precision_recall_curve

Collectives™ on Stack Overflow

AttributeError: 'Sequential' object has no attribute 'predict_proba'

2 Answers 2

Comments

2 Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

2 Comments

Linked

Related