Klaus Ackermann, Monash University - Unpacking value from a new granularly resolved global dataset on internet activity

Title: Unpacking value from a new granularly resolved global dataset on internet activity

Abstract: Sometime during 2018, an astonishing human technological milestone was reached: over half of humanity became connected to a single open communications platform, the internet. Here we show how the internet can be re-conceptualised as a data-driven insights platform at a global scale. We constructed a functional data-set covering over 1,600 cities for 7 years, from over 1.5 trillion observations. To analyse this data-set, we developed a new machine learning technique that provides accurate predictions of temporal patterns of sleep and work. We apply the same method to US city-wide electricity demand data, demonstrating improved accuracy compared to the current state of the art functional data with scalar response methods. We use this method to predict out of sample, allowing comparison of personal time use patterns worldwide.

Presenter: Klaus Ackermann, Monash University

If you would like to subscribe to the Melbourne Institute Seminar Series email list, please contact us.