Change Matters: Medication Change Prediction with Recurrent Residual Networks

Authors: Chaoqi Yang, Cao Xiao, Lucas Glass, Jimeng Sun

IJCAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluated MICRON on real inpatient and outpatient datasets. MICRON achieves 3.5% and 7.8% relative improvements over the best baseline in F1 score, respectively. We evaluate MICRON against several baselines in both inpatient and outpatient datasets.
Researcher Affiliation Collaboration Chaoqi Yang1 , Cao Xiao2 , Lucas Glass2 and Jimeng Sun1 1Department of Computer Science, University of Illinois Urbana Champaign 2Analytics Center of Excellence, IQVIA
Pseudocode No The paper describes the model architecture and algorithms using mathematical equations and descriptive text, but it does not include any explicitly labeled 'Pseudocode' or 'Algorithm' blocks.
Open Source Code Yes Codes and baselines can be found here1. 1https://github.com/ycq091044/MICRON
Open Datasets Yes We consider a benchmark inpatient dataset: MIMIC-III [Johnson et al., 2016]
Dataset Splits No The paper mentions using a 'validation set' for threshold selection ('Specifically, we load the pre-trained MICRON on the validation set'), but it does not provide specific details about the size or percentages of the train/validation/test splits.
Hardware Specification No The paper mentions training time and model size, but it does not provide specific details about the hardware used for running the experiments, such as GPU or CPU models.
Software Dependencies No The paper describes the software architecture of MICRON and mentions using Sigmoid functions and neural networks, but it does not specify any particular software libraries, frameworks, or their version numbers (e.g., PyTorch 1.x, TensorFlow 2.x).
Experiment Setup Yes The loss functions are combined by weighted sum, Ltotal = λ1L(t) rec + λ2L(t) ddi + λ3 γL(t) bce + (1 γ)L(t 1) bce + λ4 γL(t) multi + (1 γ)L(t 1) multi , where λi, i = 1, 2, 3, 4, are different weights for four types of loss functions, and γ is introduced to balance two consecutive visits. During the training, one batch contains all visits of one patient, and the loss is back-propagated after each batch. In the paper, we treat the weights as hyperparameters.