Selected article for: "high risk and maximum number"
Author: Alistair Martin; Jama Nateqi; Stefanie Gruarin; Nicolas Munsch; Isselmou Abdarahmane; Bernhard Knapp
Title: An artificial intelligence-based first-line defence against COVID-19: digitally screening citizens for risks via a chatbot
  • Document date: 2020_3_26
    • ID: 52nw9gxq_29
    Snippet: For any given set of symptoms, many possible causes could give rise to that specific presentation. We count a prediction as a true positive if the true cause is listed within the top 30 results returned by Symptoma. Note that this is the maximum number of causes returned by Symptoma for any given query. Given the possible 20,000 causes contained within Symptoma, this is the top 0.15%. Focussing on COVID-19, we can generate the following classific.....
    Document: For any given set of symptoms, many possible causes could give rise to that specific presentation. We count a prediction as a true positive if the true cause is listed within the top 30 results returned by Symptoma. Note that this is the maximum number of causes returned by Symptoma for any given query. Given the possible 20,000 causes contained within Symptoma, this is the top 0.15%. Focussing on COVID-19, we can generate the following classification: The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.25.008805 doi: bioRxiv preprint Throughout, we also assess a more strict threshold of COVID-19 being returned in the top 10 results. We refer to this stringent threshold as the "high-risk " boundary.

    Search related documents:
    Co phrase search for related documents
    • high risk and maximum number: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
    • high risk and prediction count: 1
    • high risk and specific presentation: 1, 2, 3, 4
    • high risk and stringent threshold: 1, 2
    • high risk and true cause: 1