2014
DOI: 10.1007/s10115-014-0752-0
|View full text |Cite
|
Sign up to set email alerts
|

Speeding up multiple instance learning classification rules on GPUs

Abstract: Multiple instance learning is a challenging task in supervised learning and data mining. However, algorithm performance becomes slow when learning from large-scale and high-dimensional data sets. Graphics processing units (GPUs) are being used for reducing computing time of algorithms. This paper presents an implementation of the G3P-MI algorithm on GPUs for solving multiple instance problems using classification rules. The GPU model proposed is distributable to multiple GPUs, seeking for its scalability acros… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
6
0

Year Published

2015
2015
2019
2019

Publication Types

Select...
4
2

Relationship

2
4

Authors

Journals

citations
Cited by 19 publications
(6 citation statements)
references
References 46 publications
(48 reference statements)
0
6
0
Order By: Relevance
“…Multi-instance learning, also referred to as multi-instance single-label learning, studies the problems in which an object is described by a bag of instances while associated with a single label [14], [15].…”
Section: B Multi-instance Learningmentioning
confidence: 99%
See 1 more Smart Citation
“…Multi-instance learning, also referred to as multi-instance single-label learning, studies the problems in which an object is described by a bag of instances while associated with a single label [14], [15].…”
Section: B Multi-instance Learningmentioning
confidence: 99%
“…In [12], [13], a multi-instance multi-label learning (MIML) framework was proposed for multi-label classification. In MIML, the training samples are represented as bags [14], [15], each of which is described by multiple feature vectors named instances. A bag is labeled positively if at least one of its instances is positive, while it is defined negatively if all instances in it are negative.…”
Section: Introductionmentioning
confidence: 99%
“…In Ref , they proposed an implementation for Pittsburgh classifiers, which increase the computational complexity by representing an individual as a full classifier (set of rules) rather than individual rules. Extensions of these rule‐based classifiers were proposed for multi‐instance learning . The main advantages of these proposals are their transparent scalability to multiple GPUs, since populations subsets may be assigned easily to different devices without any kind of additional overhead.…”
Section: Data Mining Tasks and Techniquesmentioning
confidence: 99%
“…GPUs are devices with multi-core architectures and massive parallel processor units, which provide fast parallel hardware for a fraction of the cost of a traditional parallel system. Since the introduction of the Computer Unified Device Architecture (CUDA) in 2007, researchers have harnessed the GPU for general purpose computing, and specifically, genetic programming [14,15], and dimensionality reduction [56].…”
Section: Implementation On Gpusmentioning
confidence: 99%
“…This process involves thousands or even millions of threads that collaborate for fast and efficient fitness computation, solving the run-time problem of the evolutionary algorithm. More specific details about the parallel implementation are out of the scope of this paper, and the reader is referred to the articles in [14,15] for GPU implementation details.…”
Section: Implementation On Gpusmentioning
confidence: 99%