Seminar by Vedran Sekara

Using public sector data to understand the human behavior (and COVID)

Vedran Sekara (ITU)

Real-time data generated by the private sector, such as mobility logs from smartphones or aggregated online search records, contains valuable information on human behavior.  In combination with environmental measurements and cutting-edge machine learning and artificial intelligence methodologies, this data can be used, for instance, to estimate detailed maps of poverty, or build computational models of disease spread. However, like any datasets passively collected through the use of technology these datasets will inevitably be biased in terms of who is represented. The most vulnerable populations, such as poor, children, and old individuals have less access to technology, and will not be equally represented. In this talk we will give examples of how wealth biases affects poverty and diseases forecast models, and also look into how measures put in place to combat COVID disproportionately affect populations.