Networking and collaboration
The group aims to increase communication, information exchange and collaboration between researchers working in clinical prediction modelling. The channels of communication are our regular Presentations and various seminar groups, where the group members and invited speakers present and discuss recent developments in the area, develop guidelines for good practice, and collaborate in various projects, including those aiming to design new analytical tools.
Hosting seminar groups for researchers
Our monthly Machine Learning and Prediction Modelling seminar group provides researchers with an opportunity to discuss new methodologies, to present projects and to discuss problems. The group meets in person at the SGDP Centre, King's IoPPN on the third Wednesday of every month at 3pm. If you are interested in joining, please contact Ewan Carr on ewan.carr@kcl.ac.uk.
Workshops
Annual workshops will be held to identify limitations of current practice in psychiatric research and to discuss and offer solutions for improvement. To watch recordings of the inaugural workshop in 2019, visit the Workshop page.
Developing guidelines for good practice
The group aims to develop best practice guidelines for prediction modelling in psychiatry. To this end we are establishing a number of working groups. Guidelines will be based on Steyerberg’s Seven steps of prediction modelling and will integrate the recommendations of the Prognosis research strategy (PROGRESS) and the guidelines of the TRIPOD statement (Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis). Working with clinicians and service users, we plan to develop strategies to successfully implement prediction models in clinical practice.
These guidelines will form the basis of short courses offered by the Department of Biostatistics and Health Informatics, King's College London and elsewhere.
Providing methodological training
The group is committed to providing high-quality training to clinical researchers and practitioners who wants to employ prediction modelling techniques to research and real-world applications. More details of such initiatives can be found at the Training in methodological skills for Prediction Modelling page.
