Dynamic Measurement Scheduling for Event Forecasting using Deep RL

Authors: Chun-Hao Chang, Mingjie Mai, Anna Goldenberg

ICML 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In our simulation, our policy outperforms heuristic-based scheduling with higher predictive gain and lower cost. In a real-world ICU mortality prediction task (MIMIC3), our policies reduce the total number of measurements by 31% or improve predictive gain by a factor of 3 as compared to physicians, under the off-policy policy evaluation.
Researcher Affiliation Academia 1University of Toronto, Toronto, ON, Canada 2Vector Institute, Toronto, ON, Canada 3The Hospital for Sick Children, Toronto, ON, Canada.
Pseudocode Yes Algorithm 1 Running policy, Algorithm 2 Generate experience for a patient at time t, Algorithm 3 Training sequential DQN
Open Source Code Yes Our data preprocessing and code are available online at https://github.com/zzzace2000/autodiagnosis.
Open Datasets Yes We then test it on MIMIC3, a real ICU temporal dataset. Johnson, A. E., Pollard, T. J., Shen, L., Li-wei, H. L., Feng, M., Ghassemi, M., Moody, B., Szolovits, P., Celi, L. A., and Mark, R. G. Mimic-iii, a freely accessible critical care database. Scientific data, 3:160035, 2016.
Dataset Splits No The paper states 'select the best performing policies based on validation set', confirming the use of a validation set. However, it does not provide specific details on the dataset splits (e.g., exact percentages or sample counts for training, validation, and test sets) for either the simulated data or MIMIC3 dataset.
Hardware Specification No The paper does not provide any specific details regarding the hardware used for running the experiments, such as GPU models, CPU types, or memory specifications.
Software Dependencies No The paper discusses various models and techniques like 'multi-layer LSTM classifier' and 'dueling deep Q-learning network (DQN)', but it does not provide specific version numbers for any software dependencies or libraries used in the implementation.
Experiment Setup Yes We list all the hyperparameters in appendix B. All the training hyperparameters are listed in Table 4.