Graph structured autoencoder

Majumdar, Angshul

doi:10.1016/j.neunet.2018.07.016

Cited by 45 publications

(9 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, as in many other fields, there has been increased usage of deep learning methods for discovering subspace clusters. Examples of deep learning subspace clustering algorithms include StructAE, which uses deep neural networks [27], SSC, a graph structured autoencoder [28], and SDEC that performs semi-supervised deep embedded clustering [29]. Kelkar et al provide a more complete overview of the wider subspace clustering field in a recent survey [30].…”

Section: Related Workmentioning

confidence: 99%

RASCL: a randomised approach to subspace clusters

Moens

Čule

Goethals

2022

Int J Data Sci Anal

View full text Add to dashboard Cite

Subspace clustering aims to discover clusters in projections of highly dimensional numerical data. In this paper, we focus on discovering small collections of highly interesting subspace clusters that do not try to cluster all data points, leaving noisy data points unclustered. To this end, we propose a randomised method that first converts the highly dimensional database to a binarised one using projected samples of the original database. Subsequently, this database is mined for frequent itemsets, which we show can be translated back to subspace clusters. In this way, we are able to explore multiple subspaces of different sizes at the same time. In our extensive experimental analysis, we show on synthetic as well as real-world data that our method is capable of discovering highly interesting subspace clusters efficiently.

show abstract

Section: Related Workmentioning

confidence: 99%

RASCL: a randomised approach to subspace clusters

Moens

Čule

Goethals

2022

Int J Data Sci Anal

View full text Add to dashboard Cite

show abstract

“…In many neural network architecture we have used an approach to avoid over fitting in a MLP-SDAE model to know version is to use dropout. In deep learning, the dropout technique way that the layer is both delivered to the hidden layer or visible layer to put off overfitting and decorate the overall performance of the model [30]. An easy technique for overall performance so is that every layer is stored withinside the network with a maintaining probability ρ unbiased of every another unit.…”

Section: The Mlp-sdae Model With Dropoutmentioning

confidence: 99%

Recognition of cancer mediating genes using the novel MLP-SDAE model

Sheet

Ghosh

et al. 2022

Preprint

View full text Add to dashboard Cite

In the present article, we introduce a deep auto-encoder model to have identified some possible genes mediating different cancer. This new version is a regression-primarily based totally fateful version primarily based totally on MLP and Stacked Denoising Auto-Encoder and the model name is MLP-SDAE. This version is educated via way of means of the usage of a Stacked Denoising Auto-Encoder for characteristic selection. Also, there’s a MLP structure for backpropagation. We have also included dropout to prevent over-fitting to improve our proposed model as well as performance. The process entails companies of gene-primarily based totally correlation coefficients and in the end decided on a few feasible genes. A specified comparative examine has been executed with a number of the prevailing deep studying procedures, such as a Recurrent Neural Network (RNN), Deep Belief Network (DBN), Deep Boltzmann Machine (DBM), Auto-encoder (AE), and Denoising Auto-encoder (DAE). Four microarray gene expression data sets like human leukemia, lung, colon, and breast cancer are 1 Springer Nature 2021 L A T E X template Recognition of cancer mediating genes using the novel MLP-SDAE model successfully applied in this model. The results are verified using the preliminary test, gene expression profile plots, biochemical pathway, t-test, and also identified GO type of the important genes based on p-values.

show abstract

“…Furthermore, researchers found the denoising model without deconvolutional layers, which is the transpose of convolutional layers [44], implies that the input and the output of the denoising model may have different sizes. To keep the size of denoised CT images equal to that of the input, U-net architecture are used in denoising 2 Computational and Mathematical Methods in Medicine LDCT images [45][46][47][48][49][50][51]. Shan et al [8] proposed the conveying path-based convolutional U-net denoising model, which is called as CPCE.…”

Section: Related Workmentioning

confidence: 99%

Low-Dose CT Image Denoising with Improving WGAN and Hybrid Loss Function

Shi

Xing

et al. 2021

Computational and Mathematical Methods in Medicine

View full text Add to dashboard Cite

The X-ray radiation from computed tomography (CT) brought us the potential risk. Simply decreasing the dose makes the CT images noisy and diagnostic performance compromised. Here, we develop a novel denoising low-dose CT image method. Our framework is based on an improved generative adversarial network coupling with the hybrid loss function, including the adversarial loss, perceptual loss, sharpness loss, and structural similarity loss. Among the loss function terms, perceptual loss and structural similarity loss are made use of to preserve textural details, and sharpness loss can make reconstruction images clear. The adversarial loss can sharp the boundary regions. The results of experiments show the proposed method can effectively remove noise and artifacts better than the state-of-the-art methods in the aspects of the visual effect, the quantitative measurements, and the texture details.

show abstract

Graph structured autoencoder

Cited by 45 publications

References 29 publications

RASCL: a randomised approach to subspace clusters

RASCL: a randomised approach to subspace clusters

Recognition of cancer mediating genes using the novel MLP-SDAE model

Low-Dose CT Image Denoising with Improving WGAN and Hybrid Loss Function

Contact Info

Product

Resources

About