Crowd Prediction at Various Public Places for Covid-19 Spread Prevention Employing Linear Regression

Abstract

Input from a simple sensor-based people counters module is taken as training data for a Linear Regression and a K-Means model for predicting crowds at various public places. The influencing factor can be either time or distance and the outcome factor be crowd strength at a place. By training a Linear Regression and a K-Means algorithm, an efficient predictive model for crowd prediction can be created.