Learning a Kernel for Multi-Task Clustering

By Quanquan Gu, Zhenhui Jessie Li, Jiawei Han. 2011

60 citations recorded by the source corpus.

Multi-task learning has received increasing attention in the past decade. Many supervised multi-task learning methods have been proposed, while unsupervised multi-task learning is still a rarely studied problem. In this paper, we propose to learn a kernel for multi-task clustering. Our goal is to learn a Reproducing Kernel Hilbert Space, in which the geometric structure of the data in each task is preserved, while the data distributions of any two tasks are as close as possible. This is formulated as a unified kernel learning framework, under which we study two types of kernel learning: nonparametric kernel learning and spectral kernel design. Both types of kernel learning can be solved by linear programming. Experiments on several cross-domain text data sets demonstrate that kernel k-means on the learned kernel can achieve better clustering results than traditional single-task clustering methods. It also outperforms the newly proposed multi-task clustering method.

Subfields: Artificial Intelligence, Computational Mechanics, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Qiang Yang
2. Yu Zhang
3. 刘晗
4. Xiaotong Zhang
5. Xianchao Zhang
6. Xiao-Lei Zhang
7. Qi Tian
8. Jiebo Luo

In Computational Mechanics

1. Qiang Yang
2. Yu Zhang
3. Xianchao Zhang
4. Xiaotong Zhang
5. 刘晗
6. Qi Tian
7. Chi Su
8. Larry S. Davis

In Computer Vision and Pattern Recognition

1. Qi Tian
2. Qiang Yang
3. Yi Yang
4. Liang Zheng
5. Yu Zhang
6. Shiliang Zhang
7. Dacheng Tao
8. Jiebo Luo

Citations in the corpus

• A Survey on Multi-Task Learning (2017, methodology): “In [128], a multi-task kernel k-means method is proposed by learning the kernel matrix via both MMD that is to minimize distances between any pair of tasks and the Laplacian regularization that helps identify a smooth kernel space.”
• On constrained spectral clustering and its applications (2012, background): “In addition, Gu et al (2011) proposed a spectral kernel design that combines multiple clustering tasks.”
• Self-Adapted Multi-Task Clustering (2016, methodology): “The algorithm in [Gu et al., 2011] learns a kernel space in which the distributions of the related tasks are close to each other and the geometric structures of the original data are preserved.”
• Smart Multi-Task Bregman Clustering and Multi-Task Kernel Clustering (2013, background): “(Gu, Li, and Han 2011) handled multitask clustering problem by learning a kernel.”
• Multi-Task Clustering with Model Relation Learning (2018, methodology): “LNKMTC and LSKMTC [Gu et al. , 2011] learn a kernel space where the task distributions are close to each other.”
• Multi-Task Learning with Low Rank Attribute Embedding for Multi-Camera Person Re-Identification (2018, methodology): “Kernel method has also been utilized to handle linearly non-separable features [55], [56].”
• Multi-Task Multi-View Clustering for Non-Negative Data (2015, methodology): “…algorithm (CoTr) [Ku- 4http://www.cs.cmu.edu/∼mccallum/bow/ mar and Daumé, 2011]; (3) multi-task clustering methods: the shared subspace learning multi-task clustering algorithm (LSSMTC) [Gu and Zhou, 2009] and the kernel learning multi-task clustering algorithm (LNKMTC) [Gu et al., 2011].”
• Convex Discriminative Multitask Clustering (2013, background): “They are 1) Learning the Shared Subspace for MTC (LSSMTC) [39], 2) Learning a Spectral Kernel for MTC (LSKMTC) [42], and 3) Multitask Bregman Clustering with Pairwise task regularization (MBC-P) [41].”
• A Survey on Multi-View Clustering (2017, unknown): “Multi-task clustering aims to improve the performance of unsupervised clustering tasks, such as [123], [124], [125], [126], [127].”
• Multi-Task Multi-View Clustering (2016, background): “The two tasks both have visual features in the images, they also have word features in the surrounding texts.”