A Multi-objective / Multi-task Learning Framework Induced by Pareto Stationarity

Authors: Michinari Momma, Chaosheng Dong, Jia Liu

ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results demonstrate not only the method achieves competitive performance with existing methods, but also it allows us to achieve the performance from different forms of preferences.
Researcher Affiliation Collaboration 1Amazon.com Inc. 2The Ohio State University.
Pseudocode Yes Algorithm 1 XWC-MGDA
Open Source Code No The paper does not provide any explicit statements about releasing source code for the described methodology or links to a code repository.
Open Datasets Yes For image classification, we use three datasets: (1) Multi MNIST (Sabour et al., 2017), (2) Multi-Fashion (Xiao et al., 2017), and (3) Multi-Fashion+MNIST (Lin et al., 2019b).
Dataset Splits No In each dataset, there are 120,000 samples in the training set and 20,000 samples in the test set. No specific information about a validation set split is provided.
Hardware Specification No The paper does not provide any specific details about the hardware used for running experiments.
Software Dependencies No The paper mentions models like 'Le Net' and 'fully connected feed-forward neural network (FNN)' but does not list specific software dependencies with version numbers (e.g., Python, PyTorch, TensorFlow versions).
Experiment Setup No The paper mentions architectural choices (Le Net, 4-layer FNN) and loss functions (MSE, SBCE), and using a random seed, but does not provide specific hyperparameter values such as learning rate, batch size, or optimizer settings for the experiments.