An incremental dimensionality reduction method on discriminant information for pattern classification

A two‐step method for predicting time to flashover in room corner test fires using cone calorimeter data

“…Compared with classifying the original data set { x }, classifying data set { y * = W * x } corresponding to the optimal projection matrix W * is relatively easier and will produce higher accuracy. The LDA technique is a very successful and effective supervised method in pattern classification . Compared with unsupervised methods, LDA is too expensive for high‐dimensional datasets because its computational complexity is determined by the dimensionality of data points.…”

“…However, as our data set has low dimensionality, it is appropriate to use the LDA for our application. The tool used for the data classification in this study is implemented in an earlier work on dimensionality reduction method for pattern classification .…”

“…However, as our data set has low dimensionality, it is appropriate to use the LDA for our application. The tool used for the data classification in this study is implemented in an earlier work on dimensionality reduction method for pattern classification [15]. Because of limited availability of data, only 52 of the materials with reliable experimental results in [3,9,16] are used to demonstrate the methodology in this study and to derive a more reliable model for predicting time to flashover in room corner fire tests.…”

Dimensionality Reduction by Weighted Connections between Neighborhoods

“…Classical MDS. MDS [4,8,16] is one of the global nonlinear techniques for dimensionality reduction which attempts to preserve global properties of the data. It maps the highdimensional data representation to a low-dimensional representation while retaining the pairwise distances between the data points as faithfully as possible.…”

“…Recently, a great many methods of dimensionality reduction methods based on manifold learning have been introduced and some classical techniques have been further improved [8][9][10][11][12][13][14][15]. In [8], Hu et al proposed a new dimensionality reduction algorithm called discriminant multidimensional mapping (DMM), which combines the advantages of multidimensional scaling (MDS) and LDA. DMM is effective for small sample datasets with high dimensionality.…”