New Wave-BLS framework enhances Broad Learning System robustness against data uncertainty

Researchers have developed a new framework called Wave-BLS, designed to make Broad Learning Systems (BLS) more robust against noisy data and corrupted labels. Unlike traditional BLS which uses squared error loss, Wave-BLS incorporates a wave loss function that offers asymmetric, bounded, and smooth penalization of errors. This new approach is optimized using a Nesterov accelerated gradient scheme, avoiding matrix inversion for better scalability. Experiments across 30 datasets showed Wave-BLS consistently outperformed standard BLS and other robust variants, demonstrating significantly slower performance degradation under controlled noise and outlier conditions. AI

IMPACT Enhances the reliability of broad learning systems in real-world applications by improving robustness to noisy data.

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

Read on arXiv cs.LG →

• arXiv
• Friedman
• Nemenyi
• Nesterov accelerated gradient
• University of California, Irvine
• Wave-BLS
• Wave Loss

• paper
• model release

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