New algorithms for online time series search with interrelated prices
Abstract
We consider the online time series search problem with interrelated prices. Using the already established UND algorithm of [5], we develop one new optimal online algorithm o-UND which improves the experimental performance of an optimal solution for selected parameter combinations. We conduct an experimental testing of UND and o-UND and establish the parameter combinations for which one algorithm is better than the other. We then combine these two algorithms into a new one called POUND. This algorithm incorporates the strengths of UND and o-UND and is also optimal.