Applied research PROBLEMS with A.I. for systematic: HF,Auto- or Algorithmic

[NSK8700]

What kind of formal computational or problems are nascent and still open for a rigorous academic research in AI (application of concepts of Artificial Intelligence) domain,as a PhD level research topic that could be empirically applied to systematic, Auto- or high-frequency trading or provide for a "Systems" (can be interpreted in a broad sense) solution with moderation of advanced mathematics.