Task sensitive feature exploration and learning for multitask graph classification

Shirui Pan, Jia Wu, Xingquan Zhu, Guodong Long, Chengqi Zhang

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Multitask learning (MTL) is commonly used for jointly optimizing multiple learning tasks. To date, all existing MTL methods have been designed for tasks with feature-vector represented instances, but cannot be applied to structure data, such as graphs. More importantly, when carrying out MTL, existing methods mainly focus on exploring overall commonality or disparity between tasks for learning, but cannot explicitly capture task relationships in the feature space, so they are unable to answer important questions, such as what exactly is shared between tasks and what is the uniqueness of one task differing from others? In this paper, we formulate a new multitask graph learning problem, and propose a task sensitive feature exploration and learning algorithm for multitask graph classification. Because graphs do not have features available, we advocate a task sensitive feature exploration and learning paradigm to jointly discover discriminative subgraph features across different tasks. In addition, a feature learning process is carried out to categorize each subgraph feature into one of three categories: 1) common feature; 2) task auxiliary feature; and 3) task specific feature, indicating whether the feature is shared by all tasks, by a subset of tasks, or by only one specific task, respectively. The feature learning and the multiple task learning are iteratively optimized to form a multitask graph classification model with a global optimization goal. Experiments on real-world functional brain analysis and chemical compound categorization demonstrate the algorithm's performance. Results confirm that our method can be used to explicitly capture task correlations and uniqueness in the feature space, and explicitly answer what are shared between tasks and what is the uniqueness of a specific task.

Original languageEnglish
Article number7430282
Pages (from-to)744-758
Number of pages15
JournalIEEE Transactions on Cybernetics
Volume47
Issue number3
DOIs
Publication statusPublished - Mar 2017
Externally publishedYes

Keywords

  • Feature selection
  • graph classification
  • multitask learning (MTL)
  • subgraph mining
  • supervised learning

Cite this

Pan, Shirui ; Wu, Jia ; Zhu, Xingquan ; Long, Guodong ; Zhang, Chengqi. / Task sensitive feature exploration and learning for multitask graph classification. In: IEEE Transactions on Cybernetics. 2017 ; Vol. 47, No. 3. pp. 744-758.
@article{bca7accb620d47fbaf50b3e440307730,
title = "Task sensitive feature exploration and learning for multitask graph classification",
abstract = "Multitask learning (MTL) is commonly used for jointly optimizing multiple learning tasks. To date, all existing MTL methods have been designed for tasks with feature-vector represented instances, but cannot be applied to structure data, such as graphs. More importantly, when carrying out MTL, existing methods mainly focus on exploring overall commonality or disparity between tasks for learning, but cannot explicitly capture task relationships in the feature space, so they are unable to answer important questions, such as what exactly is shared between tasks and what is the uniqueness of one task differing from others? In this paper, we formulate a new multitask graph learning problem, and propose a task sensitive feature exploration and learning algorithm for multitask graph classification. Because graphs do not have features available, we advocate a task sensitive feature exploration and learning paradigm to jointly discover discriminative subgraph features across different tasks. In addition, a feature learning process is carried out to categorize each subgraph feature into one of three categories: 1) common feature; 2) task auxiliary feature; and 3) task specific feature, indicating whether the feature is shared by all tasks, by a subset of tasks, or by only one specific task, respectively. The feature learning and the multiple task learning are iteratively optimized to form a multitask graph classification model with a global optimization goal. Experiments on real-world functional brain analysis and chemical compound categorization demonstrate the algorithm's performance. Results confirm that our method can be used to explicitly capture task correlations and uniqueness in the feature space, and explicitly answer what are shared between tasks and what is the uniqueness of a specific task.",
keywords = "Feature selection, graph classification, multitask learning (MTL), subgraph mining, supervised learning",
author = "Shirui Pan and Jia Wu and Xingquan Zhu and Guodong Long and Chengqi Zhang",
year = "2017",
month = "3",
doi = "10.1109/TCYB.2016.2526058",
language = "English",
volume = "47",
pages = "744--758",
journal = "IEEE Transactions on Cybernetics",
issn = "2168-2267",
number = "3",

}

Task sensitive feature exploration and learning for multitask graph classification. / Pan, Shirui; Wu, Jia; Zhu, Xingquan; Long, Guodong; Zhang, Chengqi.

In: IEEE Transactions on Cybernetics, Vol. 47, No. 3, 7430282, 03.2017, p. 744-758.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Task sensitive feature exploration and learning for multitask graph classification

AU - Pan, Shirui

AU - Wu, Jia

AU - Zhu, Xingquan

AU - Long, Guodong

AU - Zhang, Chengqi

PY - 2017/3

Y1 - 2017/3

N2 - Multitask learning (MTL) is commonly used for jointly optimizing multiple learning tasks. To date, all existing MTL methods have been designed for tasks with feature-vector represented instances, but cannot be applied to structure data, such as graphs. More importantly, when carrying out MTL, existing methods mainly focus on exploring overall commonality or disparity between tasks for learning, but cannot explicitly capture task relationships in the feature space, so they are unable to answer important questions, such as what exactly is shared between tasks and what is the uniqueness of one task differing from others? In this paper, we formulate a new multitask graph learning problem, and propose a task sensitive feature exploration and learning algorithm for multitask graph classification. Because graphs do not have features available, we advocate a task sensitive feature exploration and learning paradigm to jointly discover discriminative subgraph features across different tasks. In addition, a feature learning process is carried out to categorize each subgraph feature into one of three categories: 1) common feature; 2) task auxiliary feature; and 3) task specific feature, indicating whether the feature is shared by all tasks, by a subset of tasks, or by only one specific task, respectively. The feature learning and the multiple task learning are iteratively optimized to form a multitask graph classification model with a global optimization goal. Experiments on real-world functional brain analysis and chemical compound categorization demonstrate the algorithm's performance. Results confirm that our method can be used to explicitly capture task correlations and uniqueness in the feature space, and explicitly answer what are shared between tasks and what is the uniqueness of a specific task.

AB - Multitask learning (MTL) is commonly used for jointly optimizing multiple learning tasks. To date, all existing MTL methods have been designed for tasks with feature-vector represented instances, but cannot be applied to structure data, such as graphs. More importantly, when carrying out MTL, existing methods mainly focus on exploring overall commonality or disparity between tasks for learning, but cannot explicitly capture task relationships in the feature space, so they are unable to answer important questions, such as what exactly is shared between tasks and what is the uniqueness of one task differing from others? In this paper, we formulate a new multitask graph learning problem, and propose a task sensitive feature exploration and learning algorithm for multitask graph classification. Because graphs do not have features available, we advocate a task sensitive feature exploration and learning paradigm to jointly discover discriminative subgraph features across different tasks. In addition, a feature learning process is carried out to categorize each subgraph feature into one of three categories: 1) common feature; 2) task auxiliary feature; and 3) task specific feature, indicating whether the feature is shared by all tasks, by a subset of tasks, or by only one specific task, respectively. The feature learning and the multiple task learning are iteratively optimized to form a multitask graph classification model with a global optimization goal. Experiments on real-world functional brain analysis and chemical compound categorization demonstrate the algorithm's performance. Results confirm that our method can be used to explicitly capture task correlations and uniqueness in the feature space, and explicitly answer what are shared between tasks and what is the uniqueness of a specific task.

KW - Feature selection

KW - graph classification

KW - multitask learning (MTL)

KW - subgraph mining

KW - supervised learning

UR - http://www.scopus.com/inward/record.url?scp=84960539734&partnerID=8YFLogxK

U2 - 10.1109/TCYB.2016.2526058

DO - 10.1109/TCYB.2016.2526058

M3 - Article

VL - 47

SP - 744

EP - 758

JO - IEEE Transactions on Cybernetics

JF - IEEE Transactions on Cybernetics

SN - 2168-2267

IS - 3

M1 - 7430282

ER -