B14
Statistical physics and learning
Under what assumptions can a learning problem be expressed as a statistical-mechanical model?

IV.06 / AI
How does an error at the output determine the adjustment to a weight several layers earlier?
Backpropagation computes how a network's loss changes with its parameters by applying the chain rule through its operations. Rumelhart, Hinton, and Williams's 1986 paper gives a historical example of learning internal representations with this procedure. The practical investigation separates a correct gradient calculation from the larger questions of optimization, generalization, and what the learned representation captures.
Take the illustrative loss L(w) = (w − 3)². At w = 0, its derivative is −6. With learning rate 0.1, gradient descent updates w to 0.6; the next update gives 1.08. Compare the loss after each step, then increase the learning rate enough to observe why a valid gradient can still produce unstable optimization.
Verify the derivative of a tiny composed model numerically, then change one optimization setting and record the resulting behavior.
David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams · A neural-network learning procedure and hidden-representation experiments.
B14
Under what assumptions can a learning problem be expressed as a statistical-mechanical model?
B22
How does the distribution of variation affect the outcomes of selection and of computational search?
B24
Which feature of a neural circuit produces a measurable benefit when implemented in hardware?