Forecasting emergency medical service call arrival rates

We introduce a new method for forecasting emergency call arrival rates that combines integer-valued time series models with a dynamic latent factor structure. Covariate information is captured via simple constraints on the factor loadings. We directly model the count-valued arrivals per hour, rather than using an artificial assumption of normality. This is crucial for the emergency medical service context, in which the volume of calls may be very low. Smoothing splines are used in estimating the factor levels and loadings to improve long-term forecasts. We impose time series structure at the hourly level, rather than at the daily level, capturing the fine-scale dependence in addition to the long-term structure. Our analysis considers all emergency priority calls received by Toronto EMS between January 2007 and December 2008 for which an ambulance was dispatched. Empirical results demonstrate significantly reduced error in forecasting call arrival volume. To quantify the impact of reduced forecast errors, we design a queueing model simulation that approximates the dynamics of an ambulance system. The results show better performance as the forecasting method improves. This notion of quantifying the operational impact of improved statistical procedures may be of independent interest.

[1]  J. Stock,et al.  Macroeconomic Forecasting Using Diffusion Indexes , 2002 .

[2]  Alain Latour,et al.  Integer‐Valued GARCH Process , 2006 .

[3]  I. Priede,et al.  Improving the precision of the daily egg production method using generalized additive models , 1997 .

[4]  Yuhong Yang,et al.  Nonparametric Regression with Correlated Errors , 2001 .

[5]  Avishai Mandelbaum,et al.  Statistical Analysis of a Telephone Call Center , 2005 .

[6]  Shane G. Henderson,et al.  Should we model dependence and nonstationarity, and if so how? , 2005, Proceedings of the Winter Simulation Conference, 2005..

[7]  J. H. Schuenemeyer,et al.  Generalized Linear Models (2nd ed.) , 1992 .

[8]  Arthur J. Swersey.  Chapter 6 The deployment of police, fire, and emergency medical units , 1994, Operations research and the public sector.

[9]  Carl M. Harris,et al.  Fundamentals of queueing theory , 1975 .

[10]  S. Henderson.  Operations Research Tools for Addressing Current Challenges in Emergency Medical Services , 2011 .

[11]  Armann Ingolfsson,et al.  The application of forecasting techniques to modeling emergency medical system calls in Calgary, Alberta , 2007, Health care management science.

[12]  Eric R. Ziegel,et al.  Generalized Linear Models , 2002, Technometrics.

[13]  Haipeng Shen,et al.  Interday Forecasting and Intraday Updating of Call Center Arrivals , 2008, Manuf. Serv. Oper. Manag..

[14]  Jonathan Weinberg,et al.  Bayesian Forecasting of an Inhomogeneous Poisson Process With Applications to Call Center Data , 2007 .

[15]  Linda V. Green,et al.  INSIGHTS ON SERVICE SYSTEM DESIGN FROM A NORMAL APPROXIMATION TO ERLANG'S DELAY FORMULA , 2009 .

[16]  Sungjune Park,et al.  EMS call volume predictions: A comparative study , 2009, Comput. Oper. Res..

[17]  S. Wood.  Generalized Additive Models: An Introduction with R , 2006 .

[18]  David Vere-Jones,et al.  Point Processes , 2011, International Encyclopedia of Statistical Science.

[19]  Carl M. Harris,et al.  Fundamentals of queueing theory (2nd ed.). , 1985 .

[20]  Jeffrey E. Jarrett,et al.  Improving forecasting for telemarketing centers by ARIMA modeling with intervention , 1998 .

[21]  P. McCullagh,et al.  Generalized Linear Models , 1992 .

[22]  Chong Gu,et al.  Cross-Validating Non-Gaussian Data , 1992 .

[23]  Shane G. Henderson,et al.  Establishing Stationarity of Time Series Models via Drift Criteria , 2010 .

[24]  Henghsiu Tsai,et al.  Constrained Factor Models , 2010 .

[25]  Bruce H. Andrews,et al.  Establishing Telephone-Agent Staffing Levels through Economic Optimization , 1993 .

[26]  J. Goldberg.  Operations Research Models for the Deployment of Emergency Services Vehicles , 2004 .

[27]  Stationarity of Count-Valued and Nonlinear Time Series Models , 2010 .

[28]  Wei Lu,et al.  Constrained Independent Component Analysis , 2000, NIPS.

[29]  Yoshio Takane,et al.  Constrained Principal Component Analysis: A Comprehensive Theory , 2001, Applicable Algebra in Engineering, Communication and Computing.

[30]  Diego J. Pedregal,et al.  An unobserved component model for multi-rate forecasting of telephone call demand: the design of a forecasting support system , 2002 .

[31]  J. Michael Harrison,et al.  A Method for Staffing Large Call Centers Based on Stochastic Fluid Models , 2005, Manuf. Serv. Oper. Manag..

[32]  Alan Y. Chiang,et al.  Generalized Additive Models: An Introduction With R , 2007, Technometrics.

[33]  Michael A. West,et al.  BAYESIAN MODEL ASSESSMENT IN FACTOR ANALYSIS , 2004 .

[34]  D. Pierce,et al.  Residuals in Generalized Linear Models , 1986 .

[35]  Dag Tjøstheim,et al.  Poisson Autoregression , 2008 .

[36]  G. Daskalov,et al.  Relating fish recruitment to stock biomass and physical environment in the Black Sea using generalized additive models , 1999 .

[37]  Stanley R. Johnson,et al.  Varying Coefficient Models , 1984 .

[38]  S. Wood.  Thin plate regression splines , 2003 .

[39]  Bruce H. Andrews,et al.  L. L. Bean Improves Call-Center Forecasting , 1995 .

[40]  Haipeng Shen,et al.  FORECASTING TIME SERIES OF INHOMOGENEOUS POISSON PROCESSES WITH APPLICATION TO CALL CENTER WORKFORCE MANAGEMENT , 2008, 0807.4071.

[41]  P. McCullagh,et al.  Generalized Linear Models, 2nd Edn. , 1990 .