New Heteroskedastic Tensor Decomposition Method Introduced for EEG Data

Researchers have introduced Heteroskedastic Canonical Polyadic Tensor Decomposition (HCP), an advancement over the standard CP decomposition. HCP models entrywise variability using a non-constant, low-rank precision tensor, in addition to the low-rank mean tensor. An alternating block-coordinate ascent method has been developed to efficiently recover both tensors from noisy data, with computational complexity comparable to CP-ALS. The effectiveness of HCP has been demonstrated through synthetic experiments and an application involving electroencephalography (EEG) data. AI

IMPACT Introduces a novel statistical method for tensor decomposition, potentially improving analysis in fields like neuroscience.

RANK_REASON The item is an academic paper detailing a new statistical methodology. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv stat.ML →

• alphaXiv
• arXiv
• CatalyzeX
• CORE Recommender
• CP-ALS
• CP decomposition
• DagsHub
• electroencephalography
• Gotit.pub
• Heteroskedastic Canonical Polyadic Tensor Decomposition
• Hugging Face
• Human Connectome Project
• Influence Flower
• ScienceCast

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →