Research Area

AI for Finance

Attempts to integrate machine-learning techniques to the problems
in financial domain has been widely done in both the industry
and the academia. Among the machine-learning technologies,
we focus on deep reinforcement learning which is efficient in
solving sequential decision making problems under complex states.

For the shortest time-frame of trading, we are trying to develop
an optimal high-frequency trading policy by training the dynamics
of the limit order books. For longer time-frame of investing,
we are in progress to solve portfolio optimization under various
constraints with deep reinforcement learning.

Sample paper: "Extended Framework for Deep Reinforcement
Learning Applied to High-Frequency Trading" [working paper]


Due to the limited human resources in the financial sector, asset
management services have been offered only to the riches in the form
of private banking service. Through technology, the domain of
FinTech aims to provide these expensive financial services to
the public at low cost.

With this motivation in mind, we have developed a personalized
life-cycle goal-based investment service.
With the research as a starting point, we are currently trying to
increase the scalability of our solution, which is, to solve
the problem with lower computational cost in a faster manner.

Sample paper: "Personalized Goal-Based Investment via Multi-Stage
Stochastic Goal Programming" [link]

Investment Management

Another key research area is investment management. We aim to
develop quantitative technologies that can improve investment
performance. The main efforts have been spent on modeling
uncertainties as well as obtaining optimal investment decisions
based on such uncertainty models.

Sample paper: "Dynamic Asset Allocation for Varied Financial
Markets under Regime Switching Framework" [link]

Financial Optimization

Optimization plays a central role in financial decision making.
We study various financial optimization problems such as robust
optimization, stochastic programming, dynamic programming
from the perspective of optimal decision making under uncertainty.

Sample paper: "Deciphering Robust Portfolios" [link]


Select Presentations