Research map / Time series / Causal and generative time series
Multiple Output Regression with Latent Noise
Jussi Gillberg, Pekka Marttinen, Matti Pirinen, Antti J. Kangas, Pasi Soininen, Mehreen Ali, Aki S. Havulinna, Marjo-Riitta Järvelin, Mika Ala-Korpela, Samuel Kaski
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add latent noise model with shared parameters (latent-noise BRRR) instead of independent-noise BRRR | independent-noise BRRR · XRCC4 association detection N=4702 | detection power: 0.41 -> 0.98 | Improved |
| use latent-noise BRRR instead of exhaustive pairwise lm | lm · XRCC4 association detection N=2351 | detection power: 0.74 -> 0.94 | Improved |
| use latent-noise BRRR on a strong association gene (LIPC) | all methods · LIPC association detection | detection power: 1 -> 1.00 | No clear effect |
| add latent noise modelling instead of independent noise on DILGOM metabolomics | independent-noise BRRR · DILGOM metabolomics (predictable targets) | test MSE rel. to null: 1.04163 -> 1.00046 | Improved |
| add latent noise modelling instead of independent noise on econometrics | independent-noise BRRR · econometrics (predictable targets) | test MSE rel. to null: 0.71072 -> 0.73320 | Hurt |
| add a noise model (latent-noise BRRR) to BRRR without noise model (ablation) | BRRR without noise model · DILGOM metabolomics (predictable targets) | test MSE rel. to null: 1.30795 -> 1.00046 | Improved |
| sample the noise variance instead of cross-validating the latent signal-to-noise ratio (ablation) | latent-noise BRRR · NFBC1966 metabolomics | test MSE | Hurt |
| sample the maximum rank instead of fixing it (ablation) | latent-noise BRRR · simulations, latent noise assumption wrong | predictive performance | Improved |
| sample the rank in independent-noise BRRR instead of cross-validation (ablation) | independent-noise BRRR · NFBC1966 metabolomics | test MSE (null=1): 1.019 with N=1880 and 1.005 with N=3761 | Hurt |
| add independent noise on top of latent noise (latent+independent-noise BRRR), larger training sets (ablation) | latent-noise BRRR · NFBC1966 metabolomics N=3761 | test MSE: 0.9833 -> 0.9840 | No clear effect |
| add independent noise on top of latent noise, small econometrics set (ablation) | latent-noise BRRR · econometrics N=60 | test MSE: 0.8374 -> 0.7889 | Improved |
| use latent-noise BRRR instead of independent-noise BRRR for MCMC mixing | independent-noise BRRR · NFBC1966 MCMC convergence | averaged PSRF: 4.46 -> 1.06 | Improved |
Closest papers by meaning
- Identification of Latent Confounders via Investigating the Tensor Ranks of the Nonlinear Observations (ICML 2025 · similarity 0.85)
- Probing the Latent Hierarchical Structure of Data via Diffusion Models (ICLR 2025 · similarity 0.85)
- When Additive Noise Meets Unobserved Mediators: Bivariate Denoising Diffusion for Causal Discovery (NeurIPS 2025 · similarity 0.84)
- Kernel Regression in Structured Non-IID Settings: Theory and Implications for Denoising Score Learning (NeurIPS 2025 · similarity 0.83)
- It’s Hard to Be Normal: The Impact of Noise on Structure-agnostic Estimation (NeurIPS 2025 · similarity 0.83)
- Instrumental Variable Estimation for Causal Inference in Longitudinal Data with Time-Dependent Latent Confounders (AAAI 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
