In this video I describe the RProp training algorithm and the slight tweak to get iRProp+. We plot their solution paths in the weight-space and compare the number of iterations for the solution to both classical gradient descent and classical momentum.