Any papers support the portfolio optimization via deep learning directly?

What optimization framework can this method be applied? What the difference with supervised learning, and Deep Reinforcement Learning?

Start by reviewing the project's deep-learning portfolio-optimization scope and the questions in issue #145. Compare the requested optimization framework with supervised learning and deep reinforcement learning, and identify papers that directly support the approach. Done means providing a clear, referenced explanation addressing each question.

Written by the indexing model from the issue text.