New R-MOGP method offers outlier-robust Gaussian process modeling

Researchers have developed a new method called R-MOGP to address the sensitivity of Multi-Output Gaussian Processes (MOGPs) to outliers. This novel approach ensures robustness against anomalous data points in any individual response variable without compromising computational efficiency. R-MOGP integrates seamlessly with sparse variational inference for scalability and demonstrates performance comparable to existing state-of-the-art robust MOGPs at a significantly lower computational cost. AI

IMPACT Enhances robustness in statistical modeling for correlated data, potentially improving applications in areas like traffic prediction.

RANK_REASON The cluster contains an academic paper detailing a new statistical method. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv stat.ML →

• arXiv
• DagsHub
• Gaussian Processes
• Hugging Face
• Joshua Rooijakkers
• Multi-Output Gaussian Processes for Crowdsourced Traffic Data Imputation
• R-MOGP

• paper
• other

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