In 3rd lecture of week 3 W is described as a 4x3 vector. However I can see only 4 elements. Shouldn’t it be 4x1 vector? Can you please explain?
You might want to take a quick look back at the logistic regression video. https://www.coursera.org/learn/neural-networks-deep-learning/lecture/moUlO/vectorizing-logistic-regression and pay particular attention to the dimensions. These are for a single X ‘node’. In the Week 3 video, there are now 3 X nodes, and the shapes of the rest of the computation scale as needed.