Locality Preserving Projection via Deep Neural Network

Long, Tianhang; Gao, Junbin; Yang, Mingyan; Hu, Yongli; Yin, Baocai

doi:10.1109/ijcnn.2019.8852218

Cited by 4 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, the DNLPP method [32] is proposed, which improves LPP based on neural networks. The major difference of DNLPP and the new LPPAE method is that, although DNLPP also uses a neural network to map the input layer to the low-dimensional output layer instead of linear projection, it does not use an auto-encoder architecture, so it lacks the decoder.…”

Section: The Comparisons With the Latest Methodsmentioning

confidence: 99%

“…According to the report of DNLPP in [32], we chose to conduct experiments on MINIST, COIL-20 and ORL datasets.…”

Section: The Comparisons With the Latest Methodsmentioning

confidence: 99%

“…Recently, some new improved methods of LPP have been proposed. Based on deep neural network, a new locality preserving projection method, called DNLPP, is proposed [32]. This method employs a suitable deep neural network, rather than linear projection, to map the original sample layer and the low-dimensional embedding layer.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Locality Preserving Projections with Autoencoder

Ran¹,

Feng²,

Li³

et al. 2022

Preprint

View full text Add to dashboard Cite

<p>The proposed LPPAE method in this paper combines the conventional LPP and Autoencoder creatively. The idea of LPPAE can be made as a general framework and then extended to variants of LPP, so a series of new methods of LPP can be derived. And, the idea of LPPAE can also be extended to the other manifold-based dimensionality reduction methods, for example IsoP, NPE, LLTSA, etc., so that these methods will have new versions and achieve better performance. </p>

show abstract

Section: The Comparisons With the Latest Methodsmentioning

confidence: 99%

“…According to the report of DNLPP in [32], we chose to conduct experiments on MINIST, COIL-20 and ORL datasets.…”

Section: The Comparisons With the Latest Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Locality Preserving Projections with Autoencoder

Ran¹,

Feng²,

Li³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…LPP is a linear dimensionality reduction algorithm, 41 and its main goal is to find the transformation matrix. It solves problems that optimally preserve the neighborhood structure of the original data.…”

Section: Linear Dimensionality Reduction Techniques Taxonomymentioning

confidence: 99%

An empirical analysis of graph‐based linear dimensionality reduction techniques

Al-Omairi

Abawajy

Chowdhury

et al. 2020

Concurrency and Computation

View full text Add to dashboard Cite

Summary Many emerging applications such as social networks have prompted remarkable attention in graph data analysis. Graph data is typically high‐dimensional in nature, and dimensionality reduction is critical regarding storage, analysis, and querying of such data efficiently. Although there are many dimensionality reduction methods, it is not clear to what extent the performances of the various dimensionality reduction techniques differ. In this article, we review some of the well‐known linear dimensionality reduction methods and perform an empirical analysis of these approaches using large multidimensional graph datasets. Our results show that in linear unsupervised learning methods, the principal component analysis, singular value decomposition, and neighborhood preserving embedding methods achieve better retrieval data performance than other methods of the statistical information category, dictionary methods, and embedding methods, respectively. Regarding supervised learning methods, the experimental results demonstrate that linear discriminant analysis and partial least squares presented almost similar results.

show abstract