File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1162/neco_a_01140
- Scopus: eid_2-s2.0-85056997123
- WOS: WOS:000456919600006
- Find via
Supplementary
- Citations:
- Appears in Collections:
Article: Multi-instance dimensionality reduction via sparsity and orthogonality
Title | Multi-instance dimensionality reduction via sparsity and orthogonality |
---|---|
Authors | |
Issue Date | 2018 |
Citation | Neural Computation, 2018, v. 30, n. 12, p. 3281-3308 How to Cite? |
Abstract | © 2018 Massachusetts Institute of Technology. We study a multi-instance (MI) learning dimensionality-reduction algorithm through sparsity and orthogonality, which is especially useful for high-dimensional MI data sets. We develop a novel algorithm to handle both sparsity and orthogonality constraints that existing methods do not handle well simultaneously. Our main idea is to formulate an optimization problem where the sparse term appears in the objective function and the orthogonality term is formed as a constraint. The resulting optimization problem can be solved by using approximate augmented Lagrangian iterations as the outer loop and inertial proximal alternating linearized minimization (iPALM) iterations as the inner loop. The main advantage of this method is that both sparsity and orthogonality can be satisfied in the proposed algorithm. We show the global convergence of the proposed iterative algorithm. We also demonstrate that the proposed algorithm can achieve high sparsity and orthogonality requirements, which are very important for dimensionality reduction. Experimental results on both synthetic and real data sets show that the proposed algorithm can obtain learning performance comparable to that of other testedMI learning algorithms. |
Persistent Identifier | http://hdl.handle.net/10722/276618 |
ISSN | 2023 Impact Factor: 2.7 2023 SCImago Journal Rankings: 0.948 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Zhu, Hong | - |
dc.contributor.author | Liao, Li Zhi | - |
dc.contributor.author | Ng, Michael K. | - |
dc.date.accessioned | 2019-09-18T08:34:09Z | - |
dc.date.available | 2019-09-18T08:34:09Z | - |
dc.date.issued | 2018 | - |
dc.identifier.citation | Neural Computation, 2018, v. 30, n. 12, p. 3281-3308 | - |
dc.identifier.issn | 0899-7667 | - |
dc.identifier.uri | http://hdl.handle.net/10722/276618 | - |
dc.description.abstract | © 2018 Massachusetts Institute of Technology. We study a multi-instance (MI) learning dimensionality-reduction algorithm through sparsity and orthogonality, which is especially useful for high-dimensional MI data sets. We develop a novel algorithm to handle both sparsity and orthogonality constraints that existing methods do not handle well simultaneously. Our main idea is to formulate an optimization problem where the sparse term appears in the objective function and the orthogonality term is formed as a constraint. The resulting optimization problem can be solved by using approximate augmented Lagrangian iterations as the outer loop and inertial proximal alternating linearized minimization (iPALM) iterations as the inner loop. The main advantage of this method is that both sparsity and orthogonality can be satisfied in the proposed algorithm. We show the global convergence of the proposed iterative algorithm. We also demonstrate that the proposed algorithm can achieve high sparsity and orthogonality requirements, which are very important for dimensionality reduction. Experimental results on both synthetic and real data sets show that the proposed algorithm can obtain learning performance comparable to that of other testedMI learning algorithms. | - |
dc.language | eng | - |
dc.relation.ispartof | Neural Computation | - |
dc.title | Multi-instance dimensionality reduction via sparsity and orthogonality | - |
dc.type | Article | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1162/neco_a_01140 | - |
dc.identifier.scopus | eid_2-s2.0-85056997123 | - |
dc.identifier.volume | 30 | - |
dc.identifier.issue | 12 | - |
dc.identifier.spage | 3281 | - |
dc.identifier.epage | 3308 | - |
dc.identifier.eissn | 1530-888X | - |
dc.identifier.isi | WOS:000456919600006 | - |
dc.identifier.issnl | 0899-7667 | - |