Awesome

Probabilistic Imputation for Time-series Classification with Missing Data

Currently, we are refactoring the code. We will update README and the refactored code soon. This repository is not ready for use yet.

This repository contains the implementation for Probabilistic Imputation for Time-series Classification with Missing Data (ICML 2023).

SeungHyun Kim*, Hyunsu Kim*, EungGu Yun*, Hwangrae Lee, Jaehun Lee, Juho Lee

[Paper][ICML][BibTeX]

Installation

pip install -r requirements.txt

Datasets

Please see Set_Functions_for_Time_Series for the details and the preparation of the datasets for now. We will add the details of the datasets soon.

Usage

Train model

python scripts/train.py \
    -f, --config-file CONFIG_FILE \
    [-o, --output-dir OUTPUT_DIR] \
    [--dev] \
    [additional options]

You can change the values of the parameters in the config file using --<config-key> <value> options. For example, if you want to change the early_stopping to 10, you can use --train.early_stopping 10. Similarly, if you want to change the learning_rate to 0.1, you can use --train.learning_rate 0.1 or -lr 0.1, because -lr is registered as an alias of --train.learning_rate in scripts/train.py.

NOTE: The outputs are actually saved in outs/_/<date>-<time>-<id>/ directory. The --output-dir option just makes the link to the directory.

See help

python scripts/train.py -f CONFIG_FILE --help

Example

Basic case:
```
python scripts/train.py -f configs/physionet2012/SupNotMIWAE.yaml
```
- Use config file configs/physionet2012/SupNotMIWAE.yaml to train a SupNotMIWAE model.
- Save outputs under outs/physionet2012/SupNotMIWAE/ (default output defined in scripts/train.py)
Advanced case:
```
python scripts/train.py \
    -f configs/physionet2012/SupNotMIWAE.yaml \
    -o outs/physionet2012/SupNotMIWAE/test1 \
    -lr 0.005 \
    --model.n_units 64 \
    --dev
```
- Use config file configs/physionet2012/SupNotMIWAE.yaml to train a SupNotMIWAE model.
- Save outputs under outs/physionet2012/SupNotMIWAE/test1/.
- Set learning rate 0.005.
- Change the model arguments n_units to 64.
- Mark this as a development run.

Acknowledgement

Our code is based on Set_Functions_for_Time_Series and includes medical_ts_datasets with some modifications.

License

See LICENSE.

Citation

@inproceedings{kim2023probabilistic,
    title     = {Probabilistic Imputation for Time-series Classification with Missing Data},
    author    = {Kim, SeungHyun and Kim, Hyunsu and Yun, EungGu and Lee, Hwangrae and Lee, Jaehun and Lee, Juho},
    booktitle = {Proceedings of the 40th International Conference on Machine Learning (ICML 2023)},
    year      = {2023},
}