# Unraveling implicit human behavioral effects on dynamic characteristics of Covid-19 daily infection rates in Taiwan

## TL;DR

This study explores how human behaviors influenced the spread of Covid-19 in Taiwan by analyzing infection rate patterns across different districts and age groups.

## Contribution

The paper introduces a data-driven approach to uncover implicit behavioral effects on disease transmission dynamics at a fine scale.

## Key findings

- Distinct asymmetric growth and decline patterns were observed in infection rate curves across districts.
- Conditional entropy and mutual information identified key factors influencing peak values and curvature of infection rates.
- Behavioral effects related to living, traveling, and working were found to implicitly influence transmission dynamics.

## Abstract

We investigate the dynamic characteristics of Covid-19 daily infection rates in Taiwan during its initial surge period, focusing on 79 districts within the seven largest cities. By employing computational techniques, we extract 18 features from each district-specific curve, transforming unstructured data into structured data. Our analysis reveals distinct patterns of asymmetric growth and decline among the curves. Utilizing theoretical information measurements such as conditional entropy and mutual information, we identify major factors of order-1 and order-2 that influence the peak value and curvature at the peak of the curves, crucial features characterizing the infection rates. Additionally, we examine the impact of geographic and socioeconomic factors on the curves by encoding each of the 79 districts with two binary characteristics: North-vs-South and Urban-vs-Suburban. Furthermore, leveraging this data-driven understanding at the district level, we explore the fine-scale behavioral effects on disease spread by examining the similarity among 96 age-group-specific curves within urban districts of Taipei and suburban districts of New Taipei City, which collectively represent a substantial portion of the nation’s population. Our findings highlight the implicit influence of human behaviors related to living, traveling, and working on the dynamics of Covid-19 transmission in Taiwan.