We propose deep neural network algorithms to calculate the efficient frontier in mean–variance and mean–conditional value-at-risk (mean–CVaR) portfolio optimization problems. Starting with the ...