{"ID":2862313,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.01418","arxiv_id":"2510.01418","title":"DiffKnock: Diffusion-based Knockoff Statistics for Neural Networks Inference","abstract":"We introduce DiffKnock, a diffusion-based knockoff framework for high-dimensional feature selection with finite-sample false discovery rate (FDR) control. DiffKnock addresses two key limitations of existing knockoff methods: preserving complex feature dependencies and detecting non-linear associations. Our approach trains diffusion models to generate valid knockoffs and uses neural network--based gradient and filter statistics to construct antisymmetric feature importance measures. Through simulations, we showed that DiffKnock achieved higher power than autoencoder-based knockoffs while maintaining target FDR, indicating its superior performance in scenarios involving complex non-linear architectures. Applied to murine single-cell RNA-seq data of LPS-stimulated macrophages, DiffKnock identifies canonical NF-$κ$B target genes (Ccl3, Hmox1) and regulators (Fosb, Pdgfb). These results highlight that, by combining the flexibility of deep generative models with rigorous statistical guarantees, DiffKnock is a powerful and reliable tool for analyzing single-cell RNA-seq data, as well as high-dimensional and structured data in other domains.","short_abstract":"We introduce DiffKnock, a diffusion-based knockoff framework for high-dimensional feature selection with finite-sample false discovery rate (FDR) control. DiffKnock addresses two key limitations of existing knockoff methods: preserving complex feature dependencies and detecting non-linear associations. Our approach tra...","url_abs":"https://arxiv.org/abs/2510.01418","url_pdf":"https://arxiv.org/pdf/2510.01418v1","authors":"[\"Heng Ge\",\"Qing Lu\"]","published":"2025-10-01T19:54:23Z","proceeding":"stat.ME","tasks":"[\"stat.ME\",\"stat.AP\",\"stat.ML\"]","methods":"[\"Diffusion Model\"]","has_code":false}