Minimizing the Cross Validation Error to Mix Kernel Matrices of Heterogeneous Biological Data

Tsuda, Koji; Uda, Shinsuke; Kin, Taishin; Asai, Kiyoshi

doi:10.1023/b:nepl.0000016845.36307.d7

Cited by 7 publications

(6 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On comparison with (2), we have and on using (5) and the distributive property of the Hadamard product. This confirms that is of the form in (15), with . It is also easy to see 7 that…”

Section: B Kernel Learning Examplessupporting

confidence: 82%

See 1 more Smart Citation

Efficient Hyperkernel Learning Using Second-Order Cone Programming

Tsang

Kwok

2006

IEEE Trans. Neural Netw.

View full text Add to dashboard Cite

The kernel function plays a central role in kernel methods. Most existing methods can only adapt the kernel parameters or the kernel matrix based on empirical data. Recently, Ong et al. introduced the method of hyperkernels which can be used to learn the kernel function directly in an inductive setting. However, the associated optimization problem is a semidefinite program (SDP), which is very computationally expensive, even with the recent advances in interior point methods. In this paper, we show that this learning problem can be equivalently reformulated as a second-order cone program (SOCP), which can then be solved more efficiently than SDPs. Comparison is also made with the kernel matrix learning method proposed by Lanckriet et aL Experimental results on both classification and regression problems, with toy and real-world data sets, show that our proposed SOCP formulation has significant speedup over the original SDP formulation. Moreover, it yields better generalization than Lanckriet et al.'s method, with a speed that is comparable, or sometimes even faster, than their quadratically constrained quadratic program (QCQP) formulation.

show abstract

“…On comparison with (2), we have and on using (5) and the distributive property of the Hadamard product. This confirms that is of the form in (15), with . It is also easy to see 7 that…”

Section: B Kernel Learning Examplessupporting

confidence: 82%

“…can often be written in the form of (13) in many kernel methods, i.e., (15) where and . We also decompose in the same form as for , i.e.,…”

Section: Section Iv-bmentioning

confidence: 99%

Efficient Hyperkernel Learning Using Second-Order Cone Programming

Tsang

Kwok

2006

IEEE Trans. Neural Netw.

View full text Add to dashboard Cite

show abstract

“…In fact, the kernel matrix implicitly represents the inner product between all Paris of instances in an embedded feature space yielded by an explicit feature mapping. Since the resulting feature space may be high-dimensional or even infinite-dimensional, kernel matrix helps to have tractable and efficient computation in the original space without explicit mapping [38]. They are integrated through Log-Euclidean Mean (LogE), Arithmetic Mean (AM) and weighted-version of LogE (W-LogE).…”

Section: Kernel-based Data Fusion Methods For Gene Prioritizationmentioning

confidence: 99%

A Post-Method Condition Analysis of Using Ensemble Machine Learning for Cancer Prognosis and Diagnosis: a systematic review

Kavousi¹,

Mirsadeghi

Hosseini

et al. 2019

Preprint

View full text Add to dashboard Cite

Background Ensemble methods are supervised learning approaches that integrate different types of data or multiple individual classifiers. It has been shown that these methods can improve professional performance. Methods This study is an attempt to provide an in-depth review on 45 most relevant articles and aims to introduce 42 ensemble classifier (EC) machine learning methods used for the detection of 18 different types of cancer. Compared to other types of cancer, breast cancer, and the 22 ensemble methods introduced for its identification, is extensively investigated. The purpose of this study was to identify, map, and analyze the current academic discourse on EC machine learning methods in order to: 1. identify overarching themes emerging from empirical studies regarding EC methods, 2. determine their input data and decision-making strategies, and 3. evaluate relevant statistical procedures. Results By comparing various approaches, we can introduce Relevance Vector Machine (RVM)-based ensemble learning method that can provide optimal solutions for problems such as curse the dimensionality and high-dimensionality of feature space without missing data values. Conclusions To obtain robust performance and achieve better results, it is tactfully suggested to use multi-omics data integration, which has demonstrated to identify cancers and their subtypes more efficiently.

show abstract

“…The employment of MKL for KFDA has received much interest in the literature. Tsuda et al considered both the linear mixture and the nonlinear mixture of kernel matrices and optimized the kernel combination weights to minimize the cross-validation error for KFDA [ 14 ]. Scores of other research work have also introduced multiple kernel FDA (MK-FDA) by constructing an optimized linear combination of several base kernels under a specific constraint for mixing weights [ 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 ].…”

Section: Introductionmentioning

confidence: 99%

Feature Extraction of Electronic Nose Signals Using QPSO-Based Multiple KFDA Signal Processing

Wen

Jia

Huang

et al. 2018

Sensors

View full text Add to dashboard Cite

The aim of this research was to enhance the classification accuracy of an electronic nose (E-nose) in different detecting applications. During the learning process of the E-nose to predict the types of different odors, the prediction accuracy was not quite satisfying because the raw features extracted from sensors’ responses were regarded as the input of a classifier without any feature extraction processing. Therefore, in order to obtain more useful information and improve the E-nose’s classification accuracy, in this paper, a Weighted Kernels Fisher Discriminant Analysis (WKFDA) combined with Quantum-behaved Particle Swarm Optimization (QPSO), i.e., QWKFDA, was presented to reprocess the original feature matrix. In addition, we have also compared the proposed method with quite a few previously existing ones including Principal Component Analysis (PCA), Locality Preserving Projections (LPP), Fisher Discriminant Analysis (FDA) and Kernels Fisher Discriminant Analysis (KFDA). Experimental results proved that QWKFDA is an effective feature extraction method for E-nose in predicting the types of wound infection and inflammable gases, which shared much higher classification accuracy than those of the contrast methods.

show abstract

Minimizing the Cross Validation Error to Mix Kernel Matrices of Heterogeneous Biological Data

Cited by 7 publications

References 26 publications

Efficient Hyperkernel Learning Using Second-Order Cone Programming

Efficient Hyperkernel Learning Using Second-Order Cone Programming

A Post-Method Condition Analysis of Using Ensemble Machine Learning for Cancer Prognosis and Diagnosis: a systematic review

Feature Extraction of Electronic Nose Signals Using QPSO-Based Multiple KFDA Signal Processing

Contact Info

Product

Resources

About