Experiment Setting

The experiments are carried out on a server equipped with an Intel Xeon Platinum 8352V CPU @ 2.10GHz, featuring 16 vCPUs and 120GB of memory. The GPU used is an NVIDIA RTX 4090 with 24GB VRAM. The algorithms are implemented using PyTorch 2.5.1 and Python 3.12, with GPU acceleration supported by CUDA 12.4. The operating system is Ubuntu 22.04.

Data Hyper-Cleaning (FSBA, LFSBA, IFSBA)

To reproduce our experiment results, first generate the data by running

python -u data_cleaning.py --pretrain 0

Then run

cd ./australian_25
python -u Run_Australian.py

Then run

cd ./australian_50
python -u Run_Australian.py

Then run

cd ./breast_25
python -u Run_Breast.py

Then run

cd ./breast_50
python -u Run_Breast.py

Hyperparemeter Optimization

All results have been saved; only plotting is required.

python -u plot.py

Meta-learning Experiments

Requirements

pip install learn2learn

Running Experiments

FC100 dataset

python -u F2BAfc.py
python -u IFSBA2fc.py
python -u qNBOfc.py

miniImageNet dataset

python -u qNBOmini.py
python -u pzobo.py

Medical few-shot experiments

python -u meta_f2ba_medical.py

---

Plotting Results

python -u plotmini.py
python -u plotmini_fc100.py

Acknowledgement

The meta-learning experiments in this repository are partially built upon the implementation from:

https://github.com/sowmaster/esjacobians