Table 2.

Summary characteristics of the models participating in the Ebola Forecasting Challenge.

Team	Model description	No. Parameters	Model Type	Source
ASU	Logistic growth equation	2	Semi-mechanistic	(Pell, Kuang et al. 2016)
TOR	Phenomenological model (Incidence Decay with exponential adjustment)	3	Semi-mechanistic	(Tuite and Fisman 2016)
IMP	Stochastic transmission model with a time-varying reproductive number modeled as a random walk with a drift	2	Semi-mechanistic	(Nouvellet, Cori et al. 2017)
JMA –HHS	Stochastic SEIR model with a time-varying reproductive number modeled modeled as a multiplicative normal random walk with a log-linear drift	6	Semi-mechanistic	(Asher 2017)
McMasters-1	Generalized renewal equation	> 10	Semi-mechanistic/Hybrid	(Champredon, Li et al. 2017)
McMasters-2	Compartmental SEIR model that tracks the general community and healthcare workers with hospital and funeral transmission	27	Mechanistic/Hybrid	(Champredon, Li et al. 2017)
LSHTM	Stochastic SEIR with a random walk on transmission rate	8	Mechanistic	(Funk, Camacho et al. 2016)
CDC/NIH	Deterministic SEIR model with 3 transmission risk categories	7	Mechanistic	(Gaffey and Viboud 2017)
BI of VT	Agent-based model.	6–9, varies over time	Mechanistic	(Venkatramanan, Lewis et al. 2017)
Ensemble mean	Mean of the incidence point estimates of models 1–9	N/A	Hybrid	This paper
Ensemble BMA	Bayesian average of the incidence point estimates of models 1–9	Uninformative priors	Hybrid	This paper