an error in Recurrent Neural Network(LSTM) for tweet classification in Python

Question

I am trying to improve the result by the LSTM. in part of my project I did the following for RNN:

Below is a quick method used to train the models:

    def threshold_search(y_true, y_proba, average = None):
        best_threshold = 0
        best_score = 0
        for threshold in [i * 0.01 for i in range(100)]:
            score = f1_score(y_true=y_true, y_pred=y_proba > threshold, average=average)
            if score > best_score:
                best_threshold = threshold
                best_score = score
        search_result = {'threshold': best_threshold, 'f1': best_score}
        return search_result
    def train(model, 
              X_train, y_train, X_test, y_test, 
              checkpoint_path='model.hdf5', 
              epcohs = 25, 
              batch_size = DEFAULT_BATCH_SIZE, 
              class_weights = None, 
              fit_verbose=2,
              print_summary = True
             ):
        m = model()
        if print_summary:
            print(m.summary())
        m.fit(
            X_train, 
            y_train, 
            #this is bad practice using test data for validation, in a real case would use a seperate validation set
            validation_data=(X_test, y_test),
            epochs=epcohs, 
            batch_size=batch_size,
            class_weight=class_weights,
             #saves the most accurate model, usually you would save the one with the lowest loss
            callbacks= [
                ModelCheckpoint(checkpoint_path, monitor='val_acc', verbose=1, save_best_only=True),
                EarlyStopping(patience = 2)
            ],
            verbose=fit_verbose
        ) 
        print("\n\n****************************\n\n")
        print('Loading Best Model...')
        m.load_weights(checkpoint_path)
        predictions = m.predict(X_test, verbose=1)
        print('Validation Loss:', log_loss(y_test, predictions))
        print('Test Accuracy', (predictions.argmax(axis = 1) == y_test.argmax(axis = 1)).mean())
        print('F1 Score:', f1_score(y_test.argmax(axis = 1), predictions.argmax(axis = 1), average='weighted'))
        plot_confusion_matrix(y_test.argmax(axis = 1), predictions.argmax(axis = 1), classes=encoder.classes_)
        plt.show()    
        return m #returns best performing model

and then I used the simple implementation of an LSTM. Where The layers are as follows:

Embedding: Matrix of Word Vectors, where each vector store the "meaning" of the word. These can be trained on the fly or by existing pre-trained vector.
LSTM: RNN that allows for the "building" of state over time
Dense(64): Feed Forward Neural Network used to interpret the LSTM Output
Dense(3): This it the output of the model, 3 nodes corresponding to each class. The softmax output will ensure that the sum of values = 1.0 for each output.

def model_1():
    model = Sequential()
    model.add(Embedding(input_dim = (len(tokenizer.word_counts) + 1), output_dim = 128, input_length = MAX_SEQ_LEN))
    model.add(LSTM(128))
    model.add(Dense(64, activation='relu'))
    model.add(Dense(3, activation='softmax'))
    model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
    return model

m1 = train(model_1, 
           train_text_vec,
           y_train,
           test_text_vec,
           y_test,
           checkpoint_path='model_1.h5',
           class_weights= model.any(cws))

But I got the following output and error:

Screenshot of the error

As you can see in the screenshot, the error is:

ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

Would you please help me to solve this error?

Please post the **full** error trace (click on "4 frames" to expand them all) as *text*. — desertnaut, Sep 16 '21 at 07:58

score 0 · Answer 1 · answered Sep 16 '21 at 13:31

Based on the Keras docs and this question, class_weights expects a dictionary mapping integer class indices to weights represented by floats.

I'm not sure exactly what the line model.any(cws) is supposed to do, but typically a .any() function returns a bool or array of bools. Since class_weights is expecting a dict, it panics and throws a ValueError.

My guess is that you're confusing model weights (the numbers that make up a model) with class weights (the relative importance of the things you're trying to predict). If that's the case, leaving model_weights set to their default should fix your problem.

an error in Recurrent Neural Network(LSTM) for tweet classification in Python

1 Answers1