Online Multitask Relative Similarity Learning

Authors: Shuji Hao, Peilin Zhao, Yong Liu, Steven C. H. Hoi, Chunyan Miao

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we present our empirical studies on two real datatsets for evaluating both the efficacy and efficiency of the proposed algorithms.
Researcher Affiliation Collaboration 1Institute of High Performance Computing, A*STAR, Singapore 2Articial Intelligence Department, Ant Financial Services Group, China 3Institute for Infocomm Research, A*STAR, Singapore 4School of Information Systems, SMU, Singapore 5School of Computer Science and Engineering, NTU, Singapore
Pseudocode Yes Algorithm 1 OMTRSL: The proposed algorithm for Online Multi-Task Relative Similarity Learning. Algorithm 2 OMTRSL-Active: The proposed algorithm for Active Online Multi-Task Relative Similarity Learning.
Open Source Code No The paper mentions a link for a comparison algorithm (mt LMNN) but does not provide any concrete access or explicit statement about releasing its own source code for OMTRSL or OMTRSL-Active.
Open Datasets Yes The performances of the proposed methods are evaluated on two real-world datasets: the Isolet spoken alphabet recognition dataset [Fanty and Cole, 1990] and the news20 dataset 1. 1Available on the LIBSVM Machine Learning Repository.
Dataset Splits Yes On both datasets, we used standard 5-fold cross validation for evaluation, in which 80% of the data are used for training, and the remaining 20% are used for testing.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU/GPU models, memory) used for running its experiments. It only mentions implementation language (Matlab).
Software Dependencies No The paper states "we implement our algorithms with pure Matlab language" but does not specify a version number for Matlab or any other software dependencies with version numbers.
Experiment Setup Yes Input: Parameters C > 0 and b > 0 (for OMTRSL and OMTRSL-Active). Input: Parameters C > 0, b > 0 and δ > 0 (for OMTRSL-Active). An empirical setting of C is 1.