New DPP kernels enhance minibatch sampling for machine learning

Researchers have developed new Determinantal Point Processes (DPPs) that can improve minibatch sampling for machine learning tasks. These novel DPPs, based on wavelets, offer provably better accuracy guarantees than existing methods. The work also introduces a general technique to convert continuous DPPs into discrete kernels, which preserves variance reduction properties and enables efficient sampling for subsampling tasks like minibatch and coreset construction. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces novel sampling techniques that could improve efficiency and accuracy in training machine learning models.

RANK_REASON The cluster contains an academic paper detailing a new method for machine learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv stat.ML →

• Determinantal Point Processes
• Machine Learning
• wavelets

• paper
• other