Training ideas

Curricula, sampling schedules, pseudo-labeling and multi-task mixing, with the standard training recipe as the control.

2 ideas found

Status: Failed on benchmark Category: Training ML: Graph nn Clear all

✗ Failed on benchmark 2026

Adversarial Decision-Equivalent Training

Train a graph cost predictor not only on the nominal shortest-path decision, but on budget-limited edge perturbations that cause its predicted path to disagree with the true shortest path. The perturbation is an interdiction vector that adds known delays to selected edges, forcing the model to learn path-cost margins and relative rankings rather than merely any cost function that reproduces the nominal argmin.

Useful8/10

Difficulty6/10

Novelty6/10

Training Accuracy Generalization Stability

Paper: Decision-Focused Learning in Network Interdiction Games arXiv:2608.09036

✗ Failed on benchmark 2026

Diffusion-DPP Gradient Batches

Replace uniform minibatch sampling by a fixed-size determinantal point process whose similarity matrix is a diffusion kernel on the training-data k-NN graph. The sampler repels nearby or redundant examples while preserving multiple diffusion modes, so a small batch should cover intrinsic data geometry and provide lower-variance estimates of losses and gradients.

Useful8/10

Difficulty6/10

Novelty5/10

Training Accuracy Sample efficiency Generalization

Paper: Fast determinantal sampling on general spaces and diffusion geometry arXiv:2607.06644