Sparse Logistic Regression with Lp Penalty for Biomarker Identification

Liu, Zhenqiu; Jiang, Feng; Tian, Guo‐Liang; Wang, Suna; Sato, Fumiaki; Meltzer, Stephen J.; Tan, Ming

doi:10.2202/1544-6115.1248

Cited by 57 publications

(47 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, we evaluate the influence of smoothing, normalization, and template choice on the performance of two classification methods: penalized logistic regression (PLR) and linear SVM (hard and soft margin) when combined with a highdimensional image warping technique called symmetric normalization (SyN) (Avants et al [3]) as implemented in the Advance Normalization Tools (ANTS) software package. Penalized logistic regression (PLR) has been used before in genetics to analyze microarray and sequence data (Liu et al [34]; Park and Hastie [39]; Shevade and Keerthi [44]; Zhu and Hastie [53]) and in the context of neuroimaging applications PLR has been used before to analyze fMRI (Ryali et al [43]; Yamashita et al [52]) and regional volumes of sMRI (Casanova et al [6,7]) data. Here we use PLR to solve classification problems of very large size that result when voxels from sMRI images are used as input features.…”

Section: Introductionmentioning

confidence: 99%

Evaluating the Impact of Different Factors on Voxel-Based Classification Methods of ADNI Structural MRI Brain Images

Casanova

Maldjian

Espeland

2011

Int J Biomed Data Min

View full text Add to dashboard Cite

In this work we introduce the use of penalized logistic regression (PLR) to the problem of classification of MRI images and automatic detection of Alzheimer's disease. Classification of sMRI is approached as a large scale regularization problem which uses voxels as input features. We evaluate how differences in sMRI preprocessing steps such as smoothing, normalization, and template selection affect the performance of highdimensional classification methods. In addition, we compared the relative performance of PLR to a different approach based on support vector machines. To study these questions we used data from the Alzheimer Disease Neuroimaging Initiative (ADNI). The ADNI project follows a protocol consisting of acquisition of two images in each session, image correction steps and further evaluation by experts to obtain the optimized images. We evaluated here the impact of this optimization process on the performance of high-dimensional machine learning techniques.

show abstract

Section: Introductionmentioning

confidence: 99%

Evaluating the Impact of Different Factors on Voxel-Based Classification Methods of ADNI Structural MRI Brain Images

Casanova

Maldjian

Espeland

2011

Int J Biomed Data Min

View full text Add to dashboard Cite

show abstract

“…Filter approaches are independent of any specific learning algorithms [22,2], while wrapper approaches involve learning algorithms as part of the evaluation procedure [39,23]. Some representative wrapping methods include Sequential Floating Selection (SFS) [39], Sequential Forward Floating Selection (SFFS) [32] and sparse logistic regression based methods [27,29,14,34].…”

Section: Metaheuristics For Feature Selectionmentioning

confidence: 99%

Feature selection for high-dimensional classification using a competitive swarm optimizer

2016

View full text Add to dashboard Cite

When solving many machine learning problems such as classification, there exists a large number of input features. However, not all features are relevant for solving the problem, and sometimes, including irrelevant features may deteriorate the learning performance. Therefore, it is essential to select the most relevant features, which is known as feature selection. Many feature selection algorithms have been developed, including evolutionary algorithms or particle swarm optimization (PSO) algorithms, to find a subset of the most important features for accomplishing a particular machine learning task. However, the traditional PSO does not perform well for large scale optimization problems, which degrades the effectiveness of PSO for feature selection when the number of features dramatically increases. In this paper, we propose to use a very recent PSO variant, known as competitive swarm optimizer (CSO) that was dedicated to large-scale optimization, for solving high-dimensional feature selection problems. In addition, the CSO, which was originally developed for continuous optimization, is adapted to performing feature selection that can be considered as a combinatorial optimization problem. An archive technique is also introduced to reduce computational cost. Experiments on six benchmark datasets demonstrate that compared to the canonical PSO based and a state-of-the-art PSO variant for feature selection, the proposed CSO-based

show abstract

“…For example, sparse logistic regression has been used in the prediction of Leukemia [15], Alzheimer's disease [22] and cancers [10]. In recent years people also designed different regularization terms [13][17] [26] to a enforce more complex sparsity patterns on the learned model. However, all these works require a vector based representation of the data.…”

Section: Related Workmentioning

confidence: 99%

“…Depending on the different sparsity structures the model wants to explore, we can construct different sparsity-induced regularization terms. By adding them to the objective of conventional logistic regression we can get different types of sparse logistic regression models [13][17] [26].…”

Section: Introductionmentioning

confidence: 99%

Clinical risk prediction with multilinear sparse logistic regression

Wang

Zhang

Qian

et al. 2014

Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

Logistic regression is one core predictive modeling technique that has been used extensively in health and biomedical problems. Recently a lot of research has been focusing on enforcing sparsity on the learned model to enhance its effectiveness and interpretability, which results in sparse logistic regression model. However, no matter the original or sparse logistic regression, they require the inputs to be in vector form. This limits the applicability of logistic regression in the problems when the data cannot be naturally represented vectors (e.g., functional magnetic resonance imaging and electroencephalography signals). To handle the cases when the data are in the form of multi-dimensional arrays, we propose MulSLR: Multilinear Sparse Logistic Regression. MulSLR can be viewed as a high order extension of sparse logistic regression. Instead of solving one classification vector as in conventional logistic regression, we solve for K classification vectors in MulSLR (K is the number of modes in the data). We propose a block proximal descent approach to solve the problem and prove its convergence. The convergence rate of the proposed algorithm is also analyzed. Finally we validate the efficiency and effectiveness of Mul-SLR on predicting the onset risk of patients with Alzheimer's disease and heart failure.

show abstract

Sparse Logistic Regression with Lp Penalty for Biomarker Identification

Cited by 57 publications

References 28 publications

Evaluating the Impact of Different Factors on Voxel-Based Classification Methods of ADNI Structural MRI Brain Images

Evaluating the Impact of Different Factors on Voxel-Based Classification Methods of ADNI Structural MRI Brain Images

Feature selection for high-dimensional classification using a competitive swarm optimizer

Clinical risk prediction with multilinear sparse logistic regression

Contact Info

Product

Resources

About