Brain dead. Machine learning is hard as i dont have a very mathematical background and math shoots up sometimes.
P.S i know calculus till high school, its just that it feels so stuck

Getting familiar with matrix multiplication on the one hand and derivates with multiple variables on the other and things like back-propagation on NNs become easier.
For the derivatives just imagine treating all variables as constants while calculating the derivative, except for the one variable you actually want to derive.
Maybe this helps?