OptiSeq: Optimizing Example Ordering for In-Context Learning