How to use K-Fold cross validation with DenseNet121 model

Question

I am working on classification of images breast cancer using DensetNet121 pretrained model. I split the dataset into training, testing and validation. I want to apply k-fold cross validation. I used cross_validation from sklearn library, but I get the below error when I run the code. I tried to solve it but nothing solved the error. Anyone have idea how to solve this.

in_model = tf.keras.applications.DenseNet121(input_shape=(224,224,3),
                                            include_top=False,
                                             weights='imagenet',classes = 2)
in_model.trainable = False
inputs = tf.keras.Input(shape=(224,224,3))
x = in_model(inputs)
flat = Flatten()(x)
dense_1 = Dense(1024,activation = 'relu')(flat)
dense_2 = Dense(1024,activation = 'relu')(dense_1)
prediction = Dense(2,activation = 'softmax')(dense_2)
in_pred = Model(inputs = inputs,outputs = prediction)
validation_data=(valid_data,valid_labels)
#16
in_pred.summary()
in_pred.compile(optimizer = tf.keras.optimizers.Adagrad(learning_rate=0.0002), loss=tf.keras.losses.CategoricalCrossentropy(from_logits = False), metrics=['accuracy'])
history=in_pred.fit(train_data,train_labels,epochs = 3,batch_size=32,validation_data=validation_data)
model_result=cross_validation(in_pred, train_data, train_labels, 5)

The error:

TypeError: Cannot clone object '<keras.engine.functional.Functional object at 0x000001F82E17E3A0>'
(type <class 'keras.engine.functional.Functional'>): 
it does not seem to be a scikit-learn estimator as it does not implement a 'get_params' method.

rrossmiller · Accepted Answer · 2023-04-10T21:27:40.527

Since your model is not a scikit-learn estimator, you won't be able to use sklearn's built-in cross_validate method.

You can, however use k-fold to split your data into k-folds and get the metrics for each fold. We can use TF's built in model.evaluate, or sklearn's metrics here, too).

from sklearn.model_selection import KFold

in_model = tf.keras.applications.DenseNet121(
    input_shape=(224, 224, 3), include_top=False, weights="imagenet", classes=2
)
in_model.trainable = False
inputs = tf.keras.Input(shape=(224, 224, 3))
x = in_model(inputs)
flat = Flatten()(x)
dense_1 = Dense(1024, activation="relu")(flat)
dense_2 = Dense(1024, activation="relu")(dense_1)
prediction = Dense(2, activation="softmax")(dense_2)
in_pred = Model(inputs=inputs, outputs=prediction)
validation_data = (valid_data, valid_labels)
# 16
in_pred.summary()
in_pred.compile(
    optimizer=tf.keras.optimizers.Adagrad(learning_rate=0.0002),
    loss=tf.keras.losses.CategoricalCrossentropy(from_logits=False),
    metrics=["accuracy"],
)


kf = KFold(n_splits=10)

kf.get_n_splits(train_data)

for i, (fold_train_index, fold_+fold_test_index) in enumerate(kf.split(train_data)):
    print(f"Fold {i}:")
    print(f"  Train: index={fold_train_index}")
    print(f"  Test:  index={fold_test_index}")
    history = in_pred.fit(
        train_data[fold_train_index],
        train_labels[fold_train_index],
        epochs=3,
        batch_size=32,
        validation_data=validation_data,
    )

    in_pred.evaluate(train_data[fold_test_index],train_labels[fold_test_index])

Thank you so much for your answer. Can you please clarify for me what `test_index` and `train_index` indicate? — Eda, Apr 02 '23 at 23:46
`test_index` and `train_index` (which I changed prefix with`fold_`) indicate the indices of `train_data` that should be used for train and test data for that fold. — rrossmiller, Apr 10 '23 at 21:25

How to use K-Fold cross validation with DenseNet121 model

1 Answers1