Multi-series Time-aware Sequence Partitioning for Disease Progression Modeling

Authors: Xi Yang, Yuan Zhang, Min Chi

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

Reproducibility Variable Result LLM Response
Research Type Experimental The effectiveness of MT-TICC is first validated via a case study using a real-world hand gesture dataset with ground-truth labels. Then we further apply it for sepsis progression modeling using EHRs. The results suggest that MT-TICC can significantly outperform competitive baseline models, including the TICC.
Researcher Affiliation Academia Xi Yang , Yuan Zhang , Min Chi Department of Computer Science, North Carolina State University {yxi2, yzhang93, mchi}@ncsu.edu
Pseudocode No The paper describes the optimization steps and methodology in narrative form but does not provide structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide an explicit statement about releasing its source code or a link to a code repository for the methodology described.
Open Datasets Yes The hand gesture dataset we employed in this case study contains multichannel surface electromyographic (s EMG) signals collected from 36 participants, each of whom performed a series of hand gestures twice [Lobov et al.2019]. [Lobov et al., 2019] Sergey Lobov, Nadia Krilova, et al. EMG data for gestures Data Set, 2019. https://archive.ics.uci.edu/ml/datasets/EMG+data+for+gestures#.
Dataset Splits Yes We repeated the 5-fold cross-validation ten times and conducted a corrected paired t-test [Nadeau and Bengio2003] to compare MT-TICC and M-TICC against the TICC. All models were evaluated by repeating the 3-fold cross-validation ten times.
Hardware Specification Yes Given our entire EHRs (4,224,567 events with 14 features), it converges in 20 iterations with each iteration costing 120s (Intel i7-8700k with 32GB memory).
Software Dependencies No The paper states 'We implemented LSTM with Keras' but does not provide specific version numbers for Keras or any other software dependencies.
Experiment Setup Yes In MT-TICC, the cluster number K was 11; the window size ω was 2; the sparsity and consistency coefficients λ and β were 1e-5 and 4, respectively. for example, in MT-TICC, the window size ω was 3; the sparsity and consistency coefficients λ and β were 1e-8 and 10, respectively. We implemented LSTM with Keras and tuned the parameters by grid search.