Neural Networks Flashcards
6 cards from real Artificial Intelligence practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 6 Neural Networks flashcards as text
What is transfer learning in the context of neural networks?
Answer: Reusing a pre-trained model's weights as a starting point for a new task
Transfer learning leverages weights learned on a large dataset (e.g., ImageNet) and fine-tunes them for a different task.
Which optimizer is commonly described as using adaptive learning rates per parameter and momentum?
Answer: Adam
Adam (Adaptive Moment Estimation) combines momentum and adaptive learning rates, making it a popular default optimizer.
What is a residual connection (skip connection) in a neural network?
Answer: A shortcut that adds the input of a layer block directly to its output
Residual connections add the block's input directly to its output, enabling training of very deep networks by easing gradient flow.
What does the loss function measure in neural network training?
Answer: The discrepancy between the model's predictions and the true labels
The loss function quantifies how wrong the model's predictions are, providing the signal that drives parameter updates.
Which technique generates new training images by flipping, rotating, or cropping existing ones?
Answer: Data augmentation
Data augmentation artificially expands the training set by applying transformations, improving generalization.
What is an epoch in neural network training?
Answer: One complete pass through the entire training dataset
One epoch means the model has seen every training sample exactly once, after which weights are updated.