Digital contact tracing apps for COVID-19 need to estimate the risk that a user was infected during aparticular exposure . Machine learning methods can be used to optimize the parameters of the risk score model . This can beparticularly useful when the risk factors of the disease change, e.g., due to the evolution of new variants, or the adoption of vaccines. We show that the learning approach outperforms a strongmanually designed baseline. In particular, the parameters become harder to estimatewhen there is more missing data (e.g. due to infections which were notrecorded by the app). Nevertheless, the learning approaches outperforms the approach. Still, the learners can outperform a . strongmanual designed baseline, say the authors of this paper. In this paper we limit ourselves to simulated data, so that we can study the different factors that affect the feasibility of theapproach. Nevertheless, we show that it is difficult to estimate when there are more missingData, such as missing data, or to predict the parameters are not recorded by

Author(s) : Kevin Murphy, Abhishek Kumar, Stelios Serghiou

Links : PDF - Abstract

Code :
Coursera

Keywords : learning - risk - data - parameters - baseline -

Leave a Reply

Your email address will not be published. Required fields are marked *