Mathematics for Deep Learning — Deep Learning & Neural Networks Roadm…
The specific math that makes backpropagation and modern architectures make sense
Steps in Mathematics for Deep Learning
- Linear Algebra for Deep Learning — beginner · Vectors, matrices, tensors and the operations neural networks are built from
- Calculus & the Chain Rule — beginner · Derivatives, partial derivatives and the chain rule that backpropagation is built on
- Probability & Information Theory Basics — beginner · Entropy, cross-entropy and KL divergence — the language most loss functions are written in
- Optimization Landscapes — beginner · Convexity, local minima and saddle points in the loss surfaces neural networks navigate
- Vector Calculus for Backpropagation — beginner · Jacobians and gradients of matrix operations used when differentiating through layers
Part of
- Deep Learning & Neural Networks roadmap — the full learning path