Research map / Probabilistic and scientific modeling / Molecules and proteins
Union of Intersections (UoI) for Interpretable Data Driven Discovery and Prediction
Kristofer Bouchard, Alejandro Bujan, Fred Roosta, Shashanka Ubaru, Mr. Prabhat, Antoine Snijders, Jian-Hua Mao, Edward Chang, Michael W. Mahoney, Sharmodeep Bhattacharya
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 |
|---|---|---|---|
| separate selection and estimation using support intersections followed by averaging estimates across supports | UoI_Lasso · Synthetic linear regression | prediction accuracy | Improved |
| use UoI_Lasso instead of SCAD for functional-network estimation | Functional networks from speech-production ECoG · Speech-production ECoG recordings | selection ratio: ~ 5 × sparser | Improved |
| use UoI_Lasso instead of SCAD for functional-network estimation | Functional networks from speech-production ECoG · Speech-production ECoG recordings | relative prediction accuracy | Improved |
| use UoI_Lasso instead of other methods for phenotype prediction | Collaborative Cross mouse phenotypes · Mouse weight and rotorod performance | prediction accuracy: +1% | Improved |
| use UoI_L1Logistic instead of the base classification method | Biomedical classification tasks · Biomedical classification tasks | number of parameters: 2x-10x fewer parameters | Improved |
| use UoI_CUR instead of BasicCUR | Column subset selection · PAH and SORCH genetics data | reconstruction error | Improved |
| distribute the computations across bootstrap samples and regularization parameters | UoI_Lasso · UoI_Lasso scalability data | computational runtime: approximately two-orders of magnitude | Improved |
| increase the number of bootstraps in the selection/intersection step (ablation) | UoI_Lasso · Simulated models | false positives | Improved |
| increase the number of bootstraps in the selection/intersection step (ablation) | UoI_Lasso · Simulated models | false negatives | Hurt |
| increase the number of bootstraps in the union/estimation step (ablation) | UoI_Lasso · Simulated models | false negatives | Improved |
| increase the number of bootstraps in the union/estimation step (ablation) | UoI_Lasso · Simulated models | false positives | Hurt |
| increase the number of bootstraps in the union/estimation step (ablation) | UoI_Lasso · Simulated models | estimation stability | Improved |
Closest papers by meaning
- Approximating mutual information of high-dimensional variables using learned representations (NeurIPS 2024 · similarity 0.81)
- Training Flexible Models of Genetic Variant Effects from Functional Annotations using Accelerated Linear Algebra (ICML 2025 · similarity 0.80)
- Dropout-Based Rashomon Set Exploration for Efficient Predictive Multiplicity Estimation (ICLR 2024 · similarity 0.80)
- In Silico Mapping of Visual Categorical Selectivity Across the Whole Brain (NeurIPS 2025 · similarity 0.80)
- HORSE: Hierarchical Representation for Large-Scale Neural Subset Selection (NeurIPS 2024 · similarity 0.80)
- Functional Connectomes of Neural Networks (AAAI 2025 · similarity 0.79)
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.
