AI Paper Atlas

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

SPLICE: Fully Tractable Hierarchical Extension of ICA with Pooling

Jun-ichiro Hirayama, Aapo Hyvärinen, Motoaki Kawanabe

ICML 2017 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Tractable probabilistic models · 6 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
Add ML fine-tuning on top of layerwise SPLICE-LW (SPLICE-ML) (ablation)SPLICE two-layer · Natural Images, n=200,000test avg. log-likelihood: -221.18 -> -199.67Improved
Replace single-layer ICA with two-layer SPLICE-LWICA · Natural Images, n=100,000test avg. log-likelihood: - 270.09 -> - 222.21Improved
Replace EBM-NCE with SPLICE-MLhierarchical ICA with pooling · Natural Images, n=200,000test avg. log-likelihood: -330.04 -> -199.67Improved
Evaluate partition function numerically by AIS instead of analytically (ablation)SPLICE-ML · Natural Images, n=100,000test avg. log-likelihood: - 200.77 -> -200.76No clear effect
Increase training sample size from 100,000 to 200,000SPLICE-ML · Natural Imagestest avg. log-likelihood: - 200.77 -> -199.67No clear effect
Increase sample size from 100,000 to 200,000EBM-NCE · Natural Imagestest avg. log-likelihood: -342.12 -> -330.04Improved
Increase number of layers from 2 to 4 (extra pooling and ICA layers) (ablation)SPLICE-LW · Natural Images, n=200,000test avg. log-likelihoodImproved
Increase number of layers from 2 to 4 (ablation)SPLICE-ML · Natural Images, n=200,000test avg. log-likelihoodNo clear effect
Use SPLICE instead of EBM-NCE for second-layer source separationtwo-layer model · Synthetic complex data, Layer 2performance index / mean abs. correlationImproved
Use SPLICE instead of EBM/ICA-style estimation for first layertwo-layer model · Synthetic complex data, Layer 1, small sample sizesperformance indexImproved
Add unsupervised ML fine-tuning to SPLICE-LW (ablation)SPLICE · Synthetic dataperformance indexImproved
Use SPLICE-LW/ML instead of EBM-NCE for second-layer sourcestwo-layer model · EEG BCI IV dataset 1 (original and transfer)AUC of second-layer components for task stateImproved

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