Colloquium Speaker:
Dr. Yunhong Lyu (PhD-Stats, UoW Alumni)
Assistant Professor, Trent University,
Peterborough, Ontario
Title: Sequential Change-Point Detection in a Generalized Ornstein-Uhlenbeck Process with Cyclic Mean-Reverting Behavior
Abstract: In this talk, we present a stochastic process that is suitable for positive financial data with a cyclic mean-reverting behavior. The proposed stochastic process is a generalization of the Ornstein–Uhlenbeck process. Here are the key contributions: Firstly, within the Generalized O-U process, the mean-reverting term takes the form of a periodic function, constantly fluctuating. Our emphasis lies in detecting the change-point in drift parameters within the specialized framework of the Generalized O-U process sequentially, rather than focusing on changes in mean, variance, or covariance. Additionally, we introduce two detectors designed to identify the location of the change-point and provide the asymptotic properties of these detectors under both the null and alternative hypotheses. Finally, we perform the simulation with both generated and real data which verifies the theoretical findings of the proposed detectors.
Day & Time: Friday, October 18, 2024 at 3:00pm
Location: LT 9-118
Counts toward seminar attendance for MSc and PhD students in Math & Stats.