Demand and Supply Analytical Model for Prediction to Determine the Requirement of Ventilators during COVID-19

Abstract

In the health care sector predictive analytics is always a need but right now in times of novel coronavirus 2019 (COVID-19) pandemic it became a necessity. During the early surge of the COVID-19, around the world many hospitals are facing problems to predict the exact number of required ventilators and bed capacity especially in countries like India where mostly government healthcare units are in poor shape. Various models were implemented based on the technologies available but in regions with large populations these models are not appropriate and are unable to give accurate predictions. This paper discusses a predictive analytical model which can be used to facilitate hospital’s prediction power in determining the ventilator requirements for COVID-19 patients, using the provided data from hospitals. This paper aims on the implementation of a predictive algorithm based on the available data proportion of COVID-19 patients admitted in ICU and hospital for determination of required number of ventilators in advance. The implementation is carried out using a supervised machine learning technique called random forest regression algorithm. Moreover based on the model's prediction and the current supply of ventilators in hospitals, additional demand for ventilators could be prepared in advance for critical patients which may require it at any time. And the predicted increased demand of ventilators can be produced by the supplier industry, it is an example of industry intelligence.