Timezone: »
In healthcare, it is common for the initial goal of modeling to be the prediction of critical but rare tasks (e.g. septic shock, cardiac arrest). The reality upon deployment of such a model is often different - the goal now is to assess the risk of a patient towards these (and other) critical events. The labels and their distribution of this auxiliary goal are different from the initial labels used for training the model. Continual Learning frameworks serve as an excellent way to update a model given new data, after it has been deployed in a production environment. We introduce CLARE, a Continual Learning framework which first pre-trains on a rare task (e.g. cardiac arrest), then updates according to the labels of assessed risk, collected from the clinicians in real time - a related task. We develop a novel replay-based method to sequentially learn from new data with a different label distribution. We compare our method to a model trained in a cumulative fashion as well as one that randomly replays earlier samples it has seen. We benchmark classification architectures on a simulated dataset as well as on a clinical dataset of physiological signals.
Author Information
Bohdan Naida (University of Toronto)
Addison Weatherhead (University of Toronto)
Sana Tonekaboni (University of Toronto / Vector Institute)
Anna Goldenberg (University of Toronto)
More from the Same Authors
-
2022 : Feature Restricted Group Dropout for Robust Electronic Health Record Predictions »
Bret Nestor · Anna Goldenberg · Marzyeh Ghassemi -
2022 : Volume-based Performance not Guaranteed by Promising Patch-based Results in Medical Imaging »
Abhishek Moturu · Sayali Joshi · Andrea Doria · Anna Goldenberg -
2022 : Dissecting In-the-Wild Stress from Multimodal Sensor Data »
Sujay Nagaraj · Thomas Hartvigsen · Adrian Boch · Luca Foschini · Marzyeh Ghassemi · Sarah Goodday · Stephen Friend · Anna Goldenberg -
2022 : Modeling Heart Rate Response to Exercise with Wearables Data »
Achille Nazaret · Sana Tonekaboni · Gregory Darnell · Shirley Ren · Guillermo Sapiro · Andrew Miller -
2022 Workshop: Learning from Time Series for Health »
Sana Tonekaboni · Thomas Hartvigsen · Satya Narayan Shukla · Gunnar Rätsch · Marzyeh Ghassemi · Anna Goldenberg -
2020 Poster: What went wrong and when? Instance-wise feature importance for time-series black-box models »
Sana Tonekaboni · Shalmali Joshi · Kieran Campbell · David Duvenaud · Anna Goldenberg