Analyzing the role of model uncertainty for electronic health records

@article{Dusenberry2020AnalyzingTR,
  title={Analyzing the role of model uncertainty for electronic health records},
  author={Michael W. Dusenberry and Dustin Tran and Edward Choi and Jonas Kemp and Jeremy Nixon and Ghassen Jerfel and Katherine Heller and Andrew M. Dai},
  journal={Proceedings of the ACM Conference on Health, Inference, and Learning},
  year={2020}
}
  • Michael W. Dusenberry, Dustin Tran, +5 authors Andrew M. Dai
  • Published in CHIL 2020
  • Computer Science, Mathematics
  • Proceedings of the ACM Conference on Health, Inference, and Learning
  • In medicine, both ethical and monetary costs of incorrect predictions can be significant, and the complexity of the problems often necessitates increasingly complex models. Recent work has shown that changing just the random seed is enough for otherwise well-tuned deep neural networks to vary in their individual predicted probabilities. In light of this, we investigate the role of model uncertainty methods in the medical domain. Using RNN ensembles and various Bayesian RNNs, we show that… CONTINUE READING
    9
    Twitter Mentions

    Citations

    Publications citing this paper.
    SHOWING 1-10 OF 10 CITATIONS

    Adaptive Prediction Timing for Electronic Health Records

    VIEW 2 EXCERPTS
    CITES RESULTS & METHODS

    Federated Uncertainty-Aware Learning for Distributed Hospital EHR Data

    VIEW 1 EXCERPT
    CITES BACKGROUND

    References

    Publications referenced by this paper.
    SHOWING 1-3 OF 3 REFERENCES

    MIMIC-III, a freely accessible critical care database

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    Pattern Recognition andMachine Learning (corrected 8th printing 2009 ed.)

    • ChristopherM. Bishop
    • 2006
    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    Measuring Calibration in Deep Learning. arXiv:1904.01685 [cs, stat], April 2019

    • Jeremy Nixon, Mike Dusenberry, Linchuan Zhang, Ghassen Jerfel, Dustin Tran
    • URL http://arxiv
    • 1904
    VIEW 2 EXCERPTS
    HIGHLY INFLUENTIAL