@RossW and myself are restarting the PatientLevelPrediction workgroup. In this group we will discuss the latest methods in prediction using observational healthcare data, provide demos of the HADES tools for prediction and collaborate on exciting clinical applications for prediction.
The first call will be on November the 9th at 9am ET. We will record the meeting for those unable to join.
To join the workgroup and see information about the meetings please fill out this form: Form Here making sure to select “PLP: Patient-Level Prediction” in part 5.
We have our first workgroup meeting in 2023 this week on Wednesday 11th @ 9am ET.
During this meeting we will have an exciting presentation from @EtienneDuim on a new network methodological study using OMOP databases to examine how patient data granularity impacts on AI performances and we will also discuss 2023 OKRs for the prediction workgroup.
We have another prediction workgroup meeting tomorrow, Wednesday 8th, @ 9am ET.
During this meeting we will be demonstrating the new Strategus package and the prediction module. This is the new approach for running network studies. We will also discuss prediction tasks we can use as benchmarks.
We have another prediction workgroup meeting tomorrow, Wednesday 8th, @ 9am ET.
Kevin (Jingzhi) Yu will be presenting his exciting research idea on investigating counter factual prediction in the first part of the meeting. In the second part we will discuss prediction tasks we can use as benchmarks.
The next prediction workgroup meeting is this week, Wednesday 10th May @ 9am ET.
In this call we will have a discussion on the European and US symposiums including what people would like to see from the working group and what they will be submitting.
During this meeting we will have a group discussion on three different projects to be worked on by the group. Make sure to join if you’re interested in working on publications addressing: federated learning, investigating how our Seek Cover models performed over the past 2 years and/or the impact of metric choice for hyper-parameter tuning.
April’s PLP workgroup (April 10th @ 9am ET) call is an exciting presentation from Najia on the recent publication that compared using ATLAS/PatientLevelPrediction R package vs native R code https://www.nature.com/articles/s41598-024-52723-y