I am looking for any reference where the following problem was addressed: given the list of trades of a trader teach an AI to replicate that trader's strategy.
There are several well-known results on hedge fund replication (usually factor based) and methods focusing on the replication of the return distribution (usually by stochastic optimization) but I am looking for a more "active" replication where the machine learns to trade as a trader. The main difference is that in my case I would know all the historical trades which is usually not known with other approaches.