2025/06/14 by Hongwei Zhang, Ziqi Ye, Zhang, Hongwei +13
Computer Science · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Network Packet Processing and Optimization #Network Security and Intrusion Detection #Neural Networks and Applications
paper · pdf · doi:10.48550/arxiv.2506.12352
openalex publication_date 2025/06/14 · openalex created_date 2025/10/13 · openalex updated_date 2026/07/28
We propose Network Automatic Relevance Determination (NARD), an extension of ARD for linearly probabilistic models, to simultaneously model sparse relationships between inputs X ∈ \mathbb Rd × N and outputs Y ∈ \mathbb Rm × N, while capturing the correlation structure among the Y. NARD employs a matrix normal prior which contains a sparsity-inducing parameter to identify and discard irrelevant features, thereby promoting sparsity in the model. Algorithmically, it iteratively updates both the precision matrix and the relationship between Y and the refined inputs. To mitigate the computational inefficiencies of the \mathcal O(m3 + d3) cost per iteration, we introduce Sequential NARD, which evaluates features sequentially, and a Surrogate Function Method, leveraging an efficient approximation of the marginal likelihood and simplifying the calculation of determinant and inverse of an intermediate matrix. Combining the Sequential update with the Surrogate Function method further reduces computational costs. The computational complexity per iteration for these three methods is reduced to \mathcal O(m3+p3), \mathcal O(m3 + d2), \mathcal O(m3+p2), respectively, where p ≪ d is the final number of features in the model. Our methods demonstrate significant improvements in computational efficiency with comparable performance on both synthetic and real-world datasets.