Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Nonlinear Statistical Learning with Truncated Gaussian Graphical Models
Qinliang Su, Xuejun Liao, Changyou Chen, Lawrence Carin
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 |
|---|---|---|---|
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Boston Housing | RMSE: 2.927 -> 2.820 | Improved |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Concrete Strength | RMSE: 5.657 -> 5.395 | Improved |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Energy Efficiency | RMSE: 1.029 -> 1.244 | Hurt |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Kin8nm | RMSE: 0.088 -> 0.083 | Improved |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Naval Propulsion | RMSE: 0.00057 -> 0.003 | Hurt |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Cycle Power Plant | RMSE: 3.949 -> 4.183 | Hurt |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Protein Structure | RMSE: 4.477 -> 4.431 | Improved |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Wine Quality Red | RMSE: 0.640 -> 0.625 | Improved |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Yacht Hydrodynamic | RMSE: 0.957 -> 0.841 | Improved |
| Train TGGM with maximum-likelihood estimation instead of backpropagation | TGGM regression · Year Prediction MSD | RMSE: 8.918 -> 9.002 | Hurt |
| Use TGGM-ML instead of ReLU | classification model, 1 or 2 hidden layers as listed · MNIST test | accuracy: 98.04% -> 98.31% | Improved |
| Use TGGM-ML instead of ReLU | classification model, 1 or 2 hidden layers as listed · 20 News test | accuracy: 73.27% -> 73.74% | Improved |
Closest papers by meaning
- Unsupervised Learning with Truncated Gaussian Graphical Models (AAAI 2017 · similarity 0.89)
- A Probabilistic Framework for Nonlinearities in Stochastic Neural Networks (NeurIPS 2017 · similarity 0.87)
- Extremal graphical modeling with latent variables via convex optimization (JMLR 2025 · similarity 0.86)
- The Multiple Quantile Graphical Model (NeurIPS 2016 · similarity 0.86)
- Learning Local Neighborhoods of Non-Gaussian Graphical Models (AAAI 2025 · similarity 0.84)
- Learning Gaussian DAG Models without Condition Number Bounds (ICML 2025 · similarity 0.84)
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.
