Complementary-Label Learning

This repository gives the implementation for complementary-label learning from the ICML 2019 paper [1], the ECCV 2018 paper [2], and the NeurIPS 2017 paper [3].

Requirements

• Python 3.6
• numpy 1.14
• PyTorch 1.1
• torchvision 0.2
• Scikit-Learn 0.22

Demo

The following demo will show the results with the MNIST dataset. After running the code, you should see a text file with the results saved in the same directory. The results will have three columns: epoch number, training accuracy, and test accuracy.

## Dataset - https://www.kaggle.com/c/prudential-life-insurance-assessment/data

Download the dataset and put it inside data folder

## Install all the requirements -

pip3 install -r requirements.txt

python3 demo.py --method forward --model resnet

Methods and models

In demo.py, specify the method argument to choose one of the 5 methods available:

• ga: Gradient ascent version (Algorithm 1) in [1].
• nn: Non-negative risk estimator with the max operator in [1].
• free: Assumption-free risk estimator based on Theorem 1 in [1].
• forward: Forward correction method in [2].
• pc: Pairwise comparison with sigmoid loss in [3].

Specify the model argument:

• linear: Linear model
• mlp: Multi-layer perceptron with 2 hidden layer (128 units, 64 units)