Explain the bias-variance trade-off and how to manage it. Describe the backpropagation algorithm. What is the role of activation functions in a neural network? Explain the difference between RNN and LSTM architectures. How do you handle an imbalanced dataset? What metrics would you use to evaluate a model on such a dataset? Describe the purpose of a loss function and provide examples of different loss functions. Explain the concept of ensemble learning. How do you handle overfitting in a model? What is the difference between parametric and non-parametric models?