What is the primary purpose of 'Pooling Layers' (e.g., Max-Pooling) in Convolutional Neural Networks (CNNs)?
Pooling layers downsample feature map height and width, reducing computational overhead and parameters while retaining dominant features and translation invariance.
Which feature attribution method in Explainable AI (XAI) computes cooperative game-theoretic Shapley values to explain individual machine learning predictions?
SHAP (SHapley Additive exPlanations) is the game-theoretic method based on Shapley values. None of the listed options represents this method.
How does Simulated Annealing escape local maxima during state space optimization?
Simulated Annealing escapes local maxima by accepting downhill (worse) moves with probability $p = e^(Delta_E / T)$, which decreases as temperature T cools.
In expert system construction, what is the 'Knowledge Acquisition Bottleneck'?
Feigenbaum identified knowledge acquisition as the principal bottleneck: domain experts often use tacit heuristics that are difficult to articulate into formal if-then production rules.
What architectural innovations allow Long Short-Term Memory (LSTM) networks to capture long-range dependencies better than basic RNNs?
LSTMs use gated cell states to selectively retain, write, or discard information across long sequences, directly mitigating the vanishing gradient problem in basic RNNs.