When you found your optimizer as compared to stochastic gradient descent it had (select all that apply): Higher training accuracy Similar (+/- 5%) training accuracy Lower training accuracy Higher or Similar (+/- 5%) validation accuracy Lower validation accuracy unanswered