A Kernel Partial least square based feature selection method

Talukdar, Upasana; Hazarika, Shyamanta M.; Gan, John Q.

doi:10.1016/j.patcog.2018.05.012

Cited by 37 publications

(10 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The minimum redundancy maximum relevance (mRMR) is a variable selection algorithm that tends to select variables having a high correlation with the response variable (relevance) and the least correlation between the selected variables (redundancy) . In PLS, the variables with minimum redundancy and maximum relevancy can be found by starting with a variable having maximal mutual information with the response

y

then it greedily adds variables from PLS LWs with a maximal value of

J (W) = I (W, y) - \frac{1}{| S |} \sum_{j \in S} I (W, j),

where S is the set of already selected variables and j present a respective variable in S. Variables are ranked based on mRMR, and the top

^{'} m^{'}

number of variables are marked as influential, where

^{'} m^{'}

can be determined through validation.…”

Section: Variable Selection Methods In Plsmentioning

confidence: 99%

Comparison of variable selection methods in partial least squares regression

Mehmood

Sæbø

Liland

2020

Journal of Chemometrics

147

View full text Add to dashboard Cite

Through the remarkable progress in technology, it is getting easier and easier to generate vast amounts of variables from a given sample. The selection of variables is imperative for data reduction and for understanding the modeled relationship. Partial least squares (PLS) regression is among the modeling approaches that address high throughput data. A considerable list of variable selection methods has been introduced in PLS. Most of these methods have been reviewed in a recently conducted study. Motivated by this, we have therefore conducted a comparison of available methods for variable selection within PLS. The main focus of this study was to reveal patterns of dependencies between variable selection method and data properties, which can guide the choice of method in practical data analysis. To this aim, a simulation study was conducted with data sets having diverse properties like the number of variables, the number of samples, model complexity level, and information content. The results indicate that the above factors like the number of variables, number of samples, model complexity level, information content and variant of PLS methods, and their mutual higher‐order interactions all significantly define the prediction capabilities of the model and the choice of variable selection strategy.

show abstract

y

then it greedily adds variables from PLS LWs with a maximal value of

J (W) = I (W, y) - \frac{1}{| S |} \sum_{j \in S} I (W, j),

where S is the set of already selected variables and j present a respective variable in S. Variables are ranked based on mRMR, and the top

^{'} m^{'}

number of variables are marked as influential, where

^{'} m^{'}

can be determined through validation.…”

Section: Variable Selection Methods In Plsmentioning

confidence: 99%

Comparison of variable selection methods in partial least squares regression

Mehmood

Sæbø

Liland

2020

Journal of Chemometrics

147

View full text Add to dashboard Cite

show abstract

“…The novel fast searching algorithm that operates on sorted data sets, is developed based on the concept of straight line fitting (using least square method) [7,8] and linear regression [9]. The linear trend in the data sets is exploited as the data set is already sorted.…”

Section: Basic Ideamentioning

confidence: 99%

“…where, is the index of the ℎ element of the data set, is the ℎ data value and , are the coefficients that satisfy all data elements, is the size of the data set. The coefficients and are computed based on the curve fitting method [7,8]. The backbone of our proposed searching algorithms is the above obtained Eq.…”

Section: Basic Ideamentioning

confidence: 99%

See 1 more Smart Citation

A Novel Fast Searching Algorithm Based on Least Square Regression

Challa¹,

Chintha²,

Reddaiah³

et al. 2021

RIA

View full text Add to dashboard Cite

Currently, the machine learning group is well-understood and commonly used for predictive modelling and feature generation through linear methodologies such as reversals, principal analysis and canonical correlation analyses. All these approaches are typically intended to capture fascinating subspaces in the original space of high dimensions. These methods have all a closed-form approach because of its simple linear structures, which makes estimation and theoretical analysis for small datasets very straightforward. However, it is very common for a data set to have millions or trillions of samples and features in modern machine learning problems. We deal with the problem of fast estimation from large volumes of data for ordinary squares. The search operation is a very important operation and it is useful in many applications. Some applications when the data set size is large, the linear search takes the time which is proportional to the size of the data set. Binary search and interpolation search performs good for the search of elements in the data set in O(logn) and ⋅O(log(⋅logn)) respectively in the worst case. Now, in this paper, an effort is made to develop a novel fast searching algorithm based on the least square regression curve fitting method. The algorithm is implemented and its execution results are analyzed and compared with binary search and interpolation search performance. The proposed model is compared with the traditional methods and the proposed fast searching algorithm exhibits better performance than the traditional models.

show abstract

“…Discriminant analysis algorithms have been used for dimensionality reduction and feature extraction in many applications of computer vision [52,53]. In most discriminant analysis algorithms, the transformation matrix is found by maximizing the Fisher-Rao's criterion [54].…”

Section: Features Selection Based On Ldamentioning

confidence: 99%

An Automatic HEp-2 Specimen Analysis System Based on an Active Contours Model and an SVM Classification

2019

View full text Add to dashboard Cite

The antinuclear antibody (ANA) test is widely used for screening, diagnosing, and monitoring of autoimmune diseases. The most common methods to determine ANA are indirect immunofluorescence (IIF), performed by human epithelial type 2 (HEp-2) cells, as substrate antigen. The evaluation of ANA consist an analysis of fluorescence intensity and staining patterns. This paper presents a complete and fully automatic system able to characterize IIF images. The fluorescence intensity classification was obtained by performing an image preprocessing phase and implementing a Support Vector Machines (SVM) classifier. The cells identification problem has been addressed by developing a flexible segmentation methods, based on the Hough transform for ellipses, and on an active contours model. In order to classify the HEp-2 cells, six SVM and one k-nearest neighbors (KNN)classifiers were developed. The system was tested on a public database consisting of 2080 IIF images. Unlike almost all work presented on this topic, the proposed system automatically addresses all phases of the HEp-2 image analysis process. All results have been evaluated by comparing them with some of the most representative state-of-the-art work, demonstrating the goodness of the system in the characterization of HEp-2 images.

show abstract

A Kernel Partial least square based feature selection method

Cited by 37 publications

References 43 publications

Comparison of variable selection methods in partial least squares regression

Comparison of variable selection methods in partial least squares regression

A Novel Fast Searching Algorithm Based on Least Square Regression

An Automatic HEp-2 Specimen Analysis System Based on an Active Contours Model and an SVM Classification

Contact Info

Product

Resources

About