Uploaded February 2021 | Updated November 2026, 3 hours ago
Backpropagation on tensors and on arbitrary is a little more difficult to implement than the simple scalar version that we've seen so far. We look at what is necessary to achieve this. The payoff is that we can now build any model that we can think of, and automic differentiation do the rest.
slides: mlvu.github.io/lectures/41.DeepLearning1.annotated.pdf
lecturer: Peter Bloem
Backpropagation on tensors and on arbitrary is a little more difficult to implement than the simple scalar version that we've seen so far. We look at what is necessary to achieve this. The payoff is that we can now build any model that we can think of, and automic differentiation do the rest.
slides: mlvu.github.io/lectures/41.DeepLearning1.annotated.pdf
lecturer: Peter Bloem










