Association·arcadia
Denoising autoencoders robust to missing and corrupted data
Claim that denoising autoencoders are robust to missing and corrupted data, enable dimensionality reduction, and capture nonlinear relationships in biological systems.
Confidence
100%
active
Evidence Quote
“Denoising autoencoders (trained for accuracy despite deliberately corrupted input data) are particularly attractive for modeling biological data because they're robust to measurement noise and missing data and can capture complex relationships among parameters even with minimal data”
Relationship
Denoising autoencoder models increases Robustness to missing and corrupted data
Connections (6)
Reasoning on denoising autoencoder features and benefitsInferenceChain
Denoising autoencoders robust to missing and corrupted dataAssociation
Reasoning on G–P Atlas capturing gene-gene interactions and predictive accuracyInferenceChain
Reasoning on deep learning and autoencoder models for genomic predictionInferenceChain
Dimensionality impact and prediction improvementsInferenceChain
Modeling genotype–phenotype relationship complexityInferenceChain
Evidence
“Seminal work describing denoising autoencoders to extract robust features”
Vincent P et al. (2008). Extracting and composing robust features with denoising autoencoders doi:10.1145/1390156.1390294 ↗
“Reference describing dimensionality reduction using autoencoders”
Wang Y et al. (2016). Auto-encoder based dimensionality reduction doi:10.1016/j.neucom.2015.08.104 ↗