A Principled Approach for Learning Task Similarity in Multitask Learning

Authors: Changjian Shui, Mahdieh Abbasi, Louis-Émile Robitaille, Boyu Wang, Christian Gagné

IJCAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We assess our new algorithm empirically on several benchmarks, showing not only that we find interesting and robust task relations, but that the proposed approach outperforms the baselines, reaffirming the benefits of theoretical insight in algorithm design.
Researcher Affiliation Academia 1Universit e Laval 2University of Pennsylvania {changjian.shui.1, mahdieh.abbasi.1, louis-emile.robitaille.1}@ulaval.ca, boyuwang@seas.upenn.edu, christian.gagne@gel.ulaval.ca
Pseudocode Yes Algorithm 1 AMTNN updating algorithm
Open Source Code No The paper does not provide any explicit statement or link indicating the availability of open-source code for the described methodology.
Open Datasets Yes We first evaluate our algorithm on three benchmark datasets of digit recognition, which are datasets, MNIST, MNIST M, and SVHN. We also evaluate the proposed algorithm on Amazon reviews datasets. We extract reviews from four product categories: Books, DVD, Electronics and Kitchen appliances.
Dataset Splits No The paper mentions training and testing but does not explicitly describe a validation dataset split.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory specifications) used for running its experiments.
Software Dependencies No The paper does not specify any software dependencies with version numbers needed to replicate the experiment.
Experiment Setup No The paper describes the general network architecture (e.g., 'Le Net-5 architecture', 'two convolutional layers', 'fully connected layers') and some training techniques ('gradient reversal', 'gradient penalty') but does not provide specific hyperparameter values (e.g., learning rate, batch size, number of epochs) or detailed system-level training configurations.