Predicting IBE Webinar Registrants Using Linearization of Some Sigmoidal Functions

VINCENT, STEFANUS and Bisono, Indriati Njoto and Soewandi, Hanijanto (2024) Predicting IBE Webinar Registrants Using Linearization of Some Sigmoidal Functions. [UNSPECIFIED]

	PDF Download (1079Kb)
	PDF Download (1831Kb)

Abstract

We present a novel approach to obtain parameters of some Sigmoidal functions such as: Stevens, Gompertz, Logistics, Weibull, Brody, von Bertalaffy, & Ontogenetic. This approach extends the approach relies on iterative process to find “maximum” R2 and the traditional Ordinary Least Square (OLS) that can be implemented very easily. We demonstrate the approach in predicting IBE Webinar registrants. We also provide a new modification for the stopping rule based on Taylor approximation that can make the algorithm more robust as well as stop faster with little impact to the prediction.

Actions (login required)