A Machine Learning–Driven Information Systems Readiness Index for Predicting Strategic Cloud Adoption in Ghanaian Tertiary Education
TL;DR
This study proposes the Machine Learning-driven Information System Readiness Index (ISRI), a framework that merges Structural Equation Modeling with machine learning through the integration of SEM weights in feature engineering.
Abstract
Adopting cloud computing within the tertiary education sector is still problematic, especially in developing countries, because of the low level of readiness and the absence of predictive decision support tools. In this study, we propose the Machine Learning-driven Information System Readiness Index (ISRI) that merges Structural Equation Modeling (SEM) with machine learning through the integration of SEM weights in feature engineering. The effectiveness of this framework was tested on a dataset consisting of responses from 312 tertiary educational institutions in Ghana. The results of the experiments have shown that the Artificial Neural Network (ANN) model provided an accuracy of 93% and performed better than the baselines. An ablation study confirms the importance of SEM weighted features. Thus, the proposed ISRI framework provides a way to build a theoretically sound and computationally efficient approach that links explanatory models and predictive analytics.