Postdoc Position in Medication-related Alert Recommender Systems

The high number of alerts presented to the user can lead to alert fatigue and contribute to clinician frustration and burnout. Ignoring clinically relevant information and non-adherence to alerts can raise severe ADE. Alert Recommendation Systems (ARSs) are novel systems that can provide personalized alerts that are more relevant and better fit care providers’ expertise and needs.

You will contribute to the solution to this problem by:

1. building a proof-of-concept ARS for Amsterdam UMC's medication alerts to increase the overall alert acceptance; 
2. investigating the ARS potential value compared to the current CDSS alerts.

To achieve these goals, the candidate will explore the use and development of recommendation techniques based on machine learning, deep learning, and neurosymbolic AI to uncover potentially complex interactions between alert and patient data.

We are looking for a Postdoc candidate who is eager to contribute to better clinical decision support and safer medication prescription in the hospital settings with the help of ARS.

You will investigate how to build and evaluate a suitable ARS for Amsterdam UMC’s medication alerts. This investigation will address several methodological topics; a non-exhaustive list includes:

• Collaborative filtering;
• Content-based filtering;
• Hybrid recommender systems;
• Neurosymbolic recommender systems;
• Autoencoder-based recommender systems.

You will provide efficient and scalable implementations of your methods and will integrate them with popular open-source systems. Via the Department of Medical Informatics, you will have access to large datasets comprising millions of CDSS alerts.

Prerequisites:

• PhD in computer science, machine learning, artificial intelligence, data science, medical informatics, or related fields;
• Hands-on experience in machine learning;
• Strong programming skills in Python and machine learning libraries (e.g., PyTorch, Sklearn);
• Creative and independent mindset;
• Capable of working in multidisciplinary environments;
• Fluency in English.

Not required, but helpful:

• Experience with recommender systems;
• Experience working with real-world (medical) data, e.g., having experience with data wrangling, analysis, visualizations, etc;
• Contribution to open-source projects;
• Willingness to learn Dutch.