- Haipeng Shen(University of North Carolina at Chapel Hii)

- 2015.04.08, 10:30

- 산업경영학동(E2) Room 3221

There has been work in either OM or Statistics addressing the problem of how call centers – and other high volume service businesses – can better manage the capacity-demand mismatch that results from arrival-rate uncertainty. OM papers account for uncertainty when making staffing and scheduling decisions. Statistical models have sought to better characterize the distribution of arrival rates, by time of day, as they evolve. While each line of research has made important progresses in addressing certain elements of the problem caused by arrival-rate uncertainty, neither addresses the whole problem. We present a data-driven integrated forecasting and stochastic programming framework to cope with arrival-rate uncertainty, from call centers with single arrival streams to those with multiple dependent arrival streams. Experiments with simulated and real call-center data highlight both operational and forecasting benefits of the integrated approach.