Research map / Multi-agent systems and games / Games and equilibria
Model Linkage Selection for Cooperative Learning
Jiaying Zhou, Jie Ding, Kean Ming Tan, Vahid Tarokh
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 |
|---|---|---|---|
| use adaptive model linkage selection instead of naively pooling a contaminated learner | L1 · Breast Cancer (n=25) | prediction accuracy: 0.500 -> 0.928 | Improved |
| use adaptive linkage selection instead of L1 alone | L1 · Breast Cancer (n=25) | prediction accuracy: 0.844 -> 0.928 | Improved |
| use adaptive linkage selection instead of L1 alone | L1 · Breast Cancer (n=50) | prediction accuracy: 0.894 -> 0.952 | Improved |
| use adaptive linkage selection instead of L1 alone | KIRC regression · Kidney cancer | prediction MSE: 1.74 -> 1.69 | Improved |
| pool data from L2 (KIRP) fully rather than adaptive linkage | KIRC regression · Kidney cancer | prediction MSE: 1.69 -> 1.66 | No clear effect |
| use adaptive linkage rather than joint modeling when no true linkage | Binomial learner L1 · Epidemiological Case 3 | expected log predictive density: -8.37 -> -4.58 | Improved |
| use adaptive linkage rather than L1 alone when no true linkage | Binomial learner L1 · Epidemiological Case 3 | expected log predictive density: -4.04 -> -4.58 | Hurt |
| use adaptive linkage rather than L1 alone when linkage exists | Binomial learner L1 · Epidemiological Case 1 | expected log predictive density: -4.05 -> -3.86 | Improved |
| use adaptive linkage rather than L1 alone when linkage exists | Binomial learner L1 · Epidemiological Case 2 | expected log predictive density: -4.87 -> -4.38 | Improved |
| use adaptive linkage rather than L2 alone when linkage exists | Poisson learner L2 · Epidemiological Case 4 | expected log predictive density: -3.31 -> -3.26 | Improved |
| increase sample size per learner | greedy linkage selection · Linear regression simulation | selection accuracy: 120 | Improved |
| use greedy linkage selection instead of pooling L1 and L5 under no true linkage | logistic regression L1 · Logistic regression simulation | prediction MSE | Hurt |
Closest papers by meaning
- Collaborative Learning via Prediction Consensus (NeurIPS 2023 · similarity 0.87)
- How Distributed Collaboration Influences the Diffusion Model Training? A Theoretical Perspective (ICML 2025 · similarity 0.84)
- Cooperative Knowledge Distillation: A Learner Agnostic Approach (AAAI 2024 · similarity 0.84)
- Learning to Cooperate with Humans using Generative Agents (NeurIPS 2024 · similarity 0.84)
- Reciprocal Learning (NeurIPS 2024 · similarity 0.84)
- Model Swarms: Collaborative Search to Adapt LLM Experts via Swarm Intelligence (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.
