Grey Forecasting Models for Social Media Adoptions of Transnational Enterprises

Abstract

--The lack of adoption data posed difficulties for social media researchers and resulted in scant social media studies focusing on forecasting models. To broach the subject, this study put efforts in studying the benchmark transnational enterprises and collecting the first-hand adoption time data through the APIs of three dominating social media platforms: Youtube, Facebook, and Twitter. Grey system theory facilitates prediction problems of insufficient information, small samples, and uncertainty. Therefore, we compared the performance of two grey forecasting models, GM (1, 1) and grey Verhulst model, to gain insight into social media adoption trends. Our preliminary results indicated that GM (1, 1) performs better as a potential forecasting model to keep track of trends in adoptions of social media platforms in the market.

Day-ahead load forecasting using improved grey Verhulst model

“…This results in more homogeneous variance of the data series resulting in better modelling and improved forecast accuracy. These methods include but are not limited to logarithmic transformations, moving average and exponential transformation (Chang and Wang, 2013).…”