In very high dimensional spaces it is most likely that the gradient descent process gives us a local minimum than a saddle point of the cost function.
a. True
b. False