ResTL:

Python implementation of Transfer Learning Under Conditional Shift Based on Fuzzy Residual for regression transfer.

Getting Started

Package

• skfuzzy is necessary for fuzzy parition

pip install scikit-fuzzy

Demon

• demon_curve.py is the curve demon. Given the source model and few targe data, the predicted target model ResTL(LS) and ResTL(RD) will be given.
• demon_live.py is a live demon, you can select any number of target data with mouse.
• demon_dataset.py is the demon for Kin dataset

Citation

If you use this code for your research, please cite our papers.

@inproceedings{
    title={ Transfer Learning Under Conditional Shift Based on Fuzzy Residual},
    author={G. Chen, Y. Li, X. Liu},
    booktitle={IEEE Transactions on Cybernetics},
    year={2019},
    url={(https://ieeexplore.ieee.org/document/9097377},
}

Thanks

Thanks for the codes of TSK-FS, from Peng Xu and YuqiCui