S2E

ICML'20: Searching to Exploit Memorization Effect in Learning from Corrupted Labels (PyTorch implementation).

=======

This is the code for the paper: Searching to Exploit Memorization Effect in Learning from Corrupted Labels Quanming Yao, Hansi Yang, Bo Han, Gang Niu, James T. Kwok.

Requirements

Python = 3.7, PyTorch = 1.3.1, NumPy = 1.18.5, SciPy = 1.4.1 All packages can be installed by Conda.

Running S2E on benchmark dataset with synthetic noise (MNIST, CIFAR-10 and CIFAR-100)

Example usage for MNIST with 50% symmetric noise

python heng_mnist_main.py --noise_type symmetric --noise_rate 0.5 --num_workers 1 --n_iter 10 --n_samples 6

CIFAR-10 with 50% symmetric noise

python heng_main.py --noise_type symmetric --noise_rate 0.5 --num_workers 1 --n_iter 10 --n_samples 6

And CIFAR-100 with 50% symmetric noise

python heng_100_main.py --noise_type symmetric --noise_rate 0.5 --num_workers 1 --n_iter 10 --n_samples 6

Or see scripts (.sh files) for a quick start.

New Opportunities

• Interns, research assistants, and researcher positions are available. See requirement