My research topic deals with dynamic optimization problems (DOPs). An optimization problem can be simply stated as finding a solution which maximizes or minimizes an objective function. The objective function can be anything which takes solution as input and returns a real number. For example, the objective function can take a company's investment plan as input and output the corresponding profit. For DOPs, things get more complicated as the underlying objective function is changing over time. A general assumption in DOPs is that the underlying objective function is not changing completely randomly, but in a way which in some sense is predictable. Therefore, the difficulty for sloving DOPs not only lies in how to optimize currently, but also in how to collect historical information and how this information is utilized to aid the current optimization process.
In the meantime, I am also interested and have experience in Combinatorial Optimization, Meta-heuristics, Multi-Objective Optimization, Machine Learning and Time Series Prediction.