available · data science and machine learning roles · UKManchester, UK
bright@uzosike:~$

FHIR to OMOP Desk

Choose which simulated trusts train together and see what each one gains, with no patient record leaving a trust. Then translate a health record into the OMOP research format. All patients are synthetic.

lab / fhir-to-omop /Train togetherprod · lhrchecking APIlatency n/ano request
TrustsPOST /v1/fhir-to-omop/group
Who takes part
Training rounds320 rounds
Each round, every trust trains on its own patients and only the model weights are averaged.
Nothing trains when you click. All 31 possible groups were trained in advance, because training doesn't fit on this small server. Every score is from patients the model never saw.
$ loading trusts

Choose the trusts to see what they gain.

shell · brightuzosike.devEsc to close · every command is also a link