What are the primary metrics used in the empirical evaluation of gated recurrent neural networks (GRNNs) for sequence modeling, and how do these metrics help assess their performance?
A) Accuracy, precision, recall, and F1 score, which evaluate the model's ability to predict sequences and handle various types of errors.
B) Mean squared error, loss function convergence, and training time, which assess the model's predictive accuracy and efficiency.
C) Computational complexity, network depth, and number of parameters, which focus on the model's architecture and resource requirements.
D) Training data volume, feature extraction techniques, and hyperparameter tuning, which address data preparation and model optimization.