Because SLLE algorithm is characterized by the linear dimensionality reduction of high-dimensional complex data, it can maintain its spatial features and facilitate feature reconstruction.
In detail, the accuracy of the SLLE test is close to 90%, and the LDA and PCA + LDA tests accuracy rate is followed.
In this paper, we presented the proposed selective local linear embedding (SLLE) algorithm to reduce the dimension of collected gas data for industrial gas samples identification.
Therefore, SLLE algorithm with a low degree of freedom and simple procedure can enhance the accuracy of the selected gas samples detection, which endows higher sensitivity to olfactory machine.
Caption: Figure 5: Classification results of SLLE when (a) K = 10, (b) K = 20, and (c) K = 30.
It is worth noting that when [lambda] = 0, the SLLE is turned into the original unsupervised LLE; when [lambda] = 1, it is the supervised LLE; otherwise, it is a semisupervised LLE.
SLLE (by taking into account class label information) finds an ideal low-dimensional manifold of mapping for separating the intraclass and interclass.
Firstly, SLLE is used for reduction, mapping into the original data in a new feature space.
Laplacian eigenmaps (LE), locally linear embedding (LLE), supervised locally linear embedding (SLLE), and Spearman's rank correlation coefficient (SC2) are implemented as competing methods to compare with the proposed [SLLE-SC.sup.2] method.
To overcome this problem, Ridder  proposed a supervised LLE method (SLLE) for classification.
Additionally, SLLE can preserve distinct nonlinear features of the data of interest in a low dimensional space, which could benefit the pattern recognition of the data.
As mentioned in section 2, the SLLE was employed to eliminate redundant features.