AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Non-linear, Sparse Dimensionality Reduction via Path Lasso Penalized Autoencoders

Oskar Allerbo, Rebecka Jörnsten

JMLR 2021 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Hierarchical VAEs and posterior collapse · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use path lasso instead of parameter-wise lassosparse autoencoder · Synthetic 16 clusters (noise)explained variance R2: 0.58 -> 0.89Improved
use path lasso instead of sparse PCAsparse dimensionality reduction · Synthetic 16 clusters (noise)explained variance R2: 0.50 -> 0.89Improved
use path lasso instead of dense autoencoderautoencoder · Synthetic 16 clusters (noise)explained variance R2: 0.86 -> 0.89Improved
use path lasso instead of standard lassosparse autoencoder · Synthetic 16 clusters (no noise)explained variance R2: 0.98 -> 0.98No clear effect
use non-linear autoencoder instead of linear PCAdimensionality reduction · Synthetic 16 clusters (no noise)explained variance R2: 0.48 -> 0.98Improved
use path lasso instead of standard lassosparse autoencoder · 20 Newsgroups high sparsity (2 dims)explained variance R2: 0.015 -> 0.13Improved
use path lasso instead of sparse PCAsparse PCA · 20 Newsgroups high sparsity (4 dims)explained variance R2: 0.11 -> 0.20Improved
use path lasso instead of dense autoencoderautoencoder · 20 Newsgroups low sparsity (4 dims)explained variance R2: 0.33 -> 0.34Improved
increase latent dimensionality from 2 to 25non-linear vs linear models · 20 Newsgroupsexplained variance advantage: 0.13 -> 0.55Hurt
use path lasso instead of dense autoencoderautoencoder · AT&T faces low sparsityobservation reconstruction match: 0.66 -> 0.70Improved
use path lasso instead of standard lassosparse autoencoder · AT&T faces high sparsity10-NN reconstruction match: 0.26 -> 0.75Improved
use path lasso instead of sparse PCAsparse PCA · AT&T faces high sparsityexplained variance R2: -1.2 -> 0.57Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy