This is a nice intro to linear algebra but doesn't really get into how it's related to deep learning. A good read is Vector, Matrix, and Tensor Derivatives[0], which goes into the linear algebra needed in multi-layer neural networks, specifically during the backpropogation phase.