Bonsai - Diverse and Shallow Trees for Extreme Multi-label Classification

Research output: Contribution to journalArticle

Standard

Bonsai - Diverse and Shallow Trees for Extreme Multi-label Classification. / Khandagale, Sujay; Xiao, Han; Babbar, Rohit.

In: arXiv.org, 17.04.2019.

Research output: Contribution to journalArticle

Harvard

APA

Vancouver

Author

Bibtex - Download

@article{66a13f0c56b141edb64effdcb562ead2,
title = "Bonsai - Diverse and Shallow Trees for Extreme Multi-label Classification",
abstract = "Extreme multi-label classification (XMC) refers to supervised multi-label learning involving hundreds of thousand or even millions of labels. In this paper, we develop a suite of algorithms, called Bonsai, which generalizes the notion of label representation in XMC, and partitions the labels in the representation space to learn shallow trees. We show three concrete realizations of this label representation space including : (i) the input space which is spanned by the input features, (ii) the output space spanned by label vectors based on their co-occurrence with other labels, and (iii) the joint spaceby combining the input and output representations. Furthermore, the constraint-free multi-way partitions learnt iteratively in these spaces lead to shallow trees. By combining the effect of shallow trees and generalized label representation, Bonsai achieves the best of both worlds - fast training which is comparable to state-of-the-art tree-based methods in XMC, and much better prediction accuracy, particularly on tail-labels. On a benchmark Amazon-3M dataset with 3 million labels, Bonsai outperforms a state-of-the-art one-vs-rest method in terms of prediction accuracy, while being approximately 200 times faster to train. The code for Bonsai is available at https://github.com/xmc-aalto/bonsai.",
author = "Sujay Khandagale and Han Xiao and Rohit Babbar",
year = "2019",
month = "4",
day = "17",
language = "English",
journal = "arXiv.org",
issn = "2331-8422",

}

RIS - Download

TY - JOUR

T1 - Bonsai - Diverse and Shallow Trees for Extreme Multi-label Classification

AU - Khandagale, Sujay

AU - Xiao, Han

AU - Babbar, Rohit

PY - 2019/4/17

Y1 - 2019/4/17

N2 - Extreme multi-label classification (XMC) refers to supervised multi-label learning involving hundreds of thousand or even millions of labels. In this paper, we develop a suite of algorithms, called Bonsai, which generalizes the notion of label representation in XMC, and partitions the labels in the representation space to learn shallow trees. We show three concrete realizations of this label representation space including : (i) the input space which is spanned by the input features, (ii) the output space spanned by label vectors based on their co-occurrence with other labels, and (iii) the joint spaceby combining the input and output representations. Furthermore, the constraint-free multi-way partitions learnt iteratively in these spaces lead to shallow trees. By combining the effect of shallow trees and generalized label representation, Bonsai achieves the best of both worlds - fast training which is comparable to state-of-the-art tree-based methods in XMC, and much better prediction accuracy, particularly on tail-labels. On a benchmark Amazon-3M dataset with 3 million labels, Bonsai outperforms a state-of-the-art one-vs-rest method in terms of prediction accuracy, while being approximately 200 times faster to train. The code for Bonsai is available at https://github.com/xmc-aalto/bonsai.

AB - Extreme multi-label classification (XMC) refers to supervised multi-label learning involving hundreds of thousand or even millions of labels. In this paper, we develop a suite of algorithms, called Bonsai, which generalizes the notion of label representation in XMC, and partitions the labels in the representation space to learn shallow trees. We show three concrete realizations of this label representation space including : (i) the input space which is spanned by the input features, (ii) the output space spanned by label vectors based on their co-occurrence with other labels, and (iii) the joint spaceby combining the input and output representations. Furthermore, the constraint-free multi-way partitions learnt iteratively in these spaces lead to shallow trees. By combining the effect of shallow trees and generalized label representation, Bonsai achieves the best of both worlds - fast training which is comparable to state-of-the-art tree-based methods in XMC, and much better prediction accuracy, particularly on tail-labels. On a benchmark Amazon-3M dataset with 3 million labels, Bonsai outperforms a state-of-the-art one-vs-rest method in terms of prediction accuracy, while being approximately 200 times faster to train. The code for Bonsai is available at https://github.com/xmc-aalto/bonsai.

UR - https://arxiv.org/abs/1904.08249

M3 - Article

JO - arXiv.org

JF - arXiv.org

SN - 2331-8422

ER -

ID: 36516478