Abstract
Integrating multi-source information has recently shown promising performance in predicting Alzheimer's disease (AD). Multiple kernel learning (MKL) plays an important role in this regard by learning the combination weights of a set of base kernels via the principle of margin maximisation. The latest research on MKL further incorporates the radius of minimum enclosing ball (MEB) of training data to improve the kernel learning performance. However, we observe that directly applying these radius-incorporated MKL algorithms to AD prediction tasks does not necessarily improve, and sometimes even deteriorate, the prediction accuracy. In this paper, we propose an improved radius-incorporated MKL algorithm for AD prediction. First, we redesign the objective function by approximating the radius of MEB with its upper bound, a linear function of the kernel weights. This approximation makes the resulting optimisation problem convex and globally solvable. Second, instead of using cross-validation, we model the regularisation parameter C of the SVM classifier as an extra kernel weight and automatically tune it in MKL. Third, we theoretically show that our algorithm can be reformulated into a similar form of the SimpleMKL algorithm and conveniently solved by the off-the-shelf packages. We discuss the factors that contribute to the improved performance and apply our algorithm to discriminate different clinic groups from the benchmark ADNI data set. As experimentally demonstrated, our algorithm can better utilise the radius information and achieve higher prediction accuracy than the comparable MKL methods in the literature. In addition, our algorithm demonstrates the highest computational efficiency among all the comparable methods.
| Original language | English |
|---|---|
| Pages (from-to) | 2141-2150 |
| Number of pages | 10 |
| Journal | Pattern Recognition |
| Volume | 48 |
| Issue number | 7 |
| DOIs | |
| Publication status | Published - 2015 Jul 1 |
Fingerprint
Keywords
- Alzheimer's disease
- Multiple kernel learning
- Neuroimaging
- Radius-margin bound
- Support vector machines
ASJC Scopus subject areas
- Software
- Artificial Intelligence
- Computer Vision and Pattern Recognition
- Signal Processing
Cite this
An efficient radius-incorporated MKL algorithm for Alzheimer's disease prediction. / Liu, Xinwang; Zhou, Luping; Wang, Lei; Zhang, Jian; Yin, Jianping; Shen, Dinggang.
In: Pattern Recognition, Vol. 48, No. 7, 01.07.2015, p. 2141-2150.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - An efficient radius-incorporated MKL algorithm for Alzheimer's disease prediction
AU - Liu, Xinwang
AU - Zhou, Luping
AU - Wang, Lei
AU - Zhang, Jian
AU - Yin, Jianping
AU - Shen, Dinggang
PY - 2015/7/1
Y1 - 2015/7/1
N2 - Integrating multi-source information has recently shown promising performance in predicting Alzheimer's disease (AD). Multiple kernel learning (MKL) plays an important role in this regard by learning the combination weights of a set of base kernels via the principle of margin maximisation. The latest research on MKL further incorporates the radius of minimum enclosing ball (MEB) of training data to improve the kernel learning performance. However, we observe that directly applying these radius-incorporated MKL algorithms to AD prediction tasks does not necessarily improve, and sometimes even deteriorate, the prediction accuracy. In this paper, we propose an improved radius-incorporated MKL algorithm for AD prediction. First, we redesign the objective function by approximating the radius of MEB with its upper bound, a linear function of the kernel weights. This approximation makes the resulting optimisation problem convex and globally solvable. Second, instead of using cross-validation, we model the regularisation parameter C of the SVM classifier as an extra kernel weight and automatically tune it in MKL. Third, we theoretically show that our algorithm can be reformulated into a similar form of the SimpleMKL algorithm and conveniently solved by the off-the-shelf packages. We discuss the factors that contribute to the improved performance and apply our algorithm to discriminate different clinic groups from the benchmark ADNI data set. As experimentally demonstrated, our algorithm can better utilise the radius information and achieve higher prediction accuracy than the comparable MKL methods in the literature. In addition, our algorithm demonstrates the highest computational efficiency among all the comparable methods.
AB - Integrating multi-source information has recently shown promising performance in predicting Alzheimer's disease (AD). Multiple kernel learning (MKL) plays an important role in this regard by learning the combination weights of a set of base kernels via the principle of margin maximisation. The latest research on MKL further incorporates the radius of minimum enclosing ball (MEB) of training data to improve the kernel learning performance. However, we observe that directly applying these radius-incorporated MKL algorithms to AD prediction tasks does not necessarily improve, and sometimes even deteriorate, the prediction accuracy. In this paper, we propose an improved radius-incorporated MKL algorithm for AD prediction. First, we redesign the objective function by approximating the radius of MEB with its upper bound, a linear function of the kernel weights. This approximation makes the resulting optimisation problem convex and globally solvable. Second, instead of using cross-validation, we model the regularisation parameter C of the SVM classifier as an extra kernel weight and automatically tune it in MKL. Third, we theoretically show that our algorithm can be reformulated into a similar form of the SimpleMKL algorithm and conveniently solved by the off-the-shelf packages. We discuss the factors that contribute to the improved performance and apply our algorithm to discriminate different clinic groups from the benchmark ADNI data set. As experimentally demonstrated, our algorithm can better utilise the radius information and achieve higher prediction accuracy than the comparable MKL methods in the literature. In addition, our algorithm demonstrates the highest computational efficiency among all the comparable methods.
KW - Alzheimer's disease
KW - Multiple kernel learning
KW - Neuroimaging
KW - Radius-margin bound
KW - Support vector machines
UR - http://www.scopus.com/inward/record.url?scp=84925682671&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84925682671&partnerID=8YFLogxK
U2 - 10.1016/j.patcog.2014.12.007
DO - 10.1016/j.patcog.2014.12.007
M3 - Article
AN - SCOPUS:84925682671
VL - 48
SP - 2141
EP - 2150
JO - Pattern Recognition
JF - Pattern Recognition
SN - 0031-3203
IS - 7
ER -