variance reduction
Techniques used in machine learning to lower the variability of model estimates, enhancing stability and accuracy, such as pooling or using ensemble methods.
- Computation and Memory-Efficient Model Compression with Gradient Reweighting
- Energy-based generator matching: A neural sampler for general state space
- Neural MJD: Neural Non-Stationary Merton Jump Diffusion for Time Series Prediction
- On the Convergence of Stochastic Smoothed Multi-Level Compositional Gradient Descent Ascent
- Pass@K Policy Optimization: Solving Harder Reinforcement Learning Problems
- Reverse-Annealed Sequential Monte Carlo for Efficient Bayesian Optimal Experiment Design
- STITCH-OPE: Trajectory Stitching with Guided Diffusion for Off-Policy Evaluation
- Subsampled Ensemble Can Improve Generalization Tail Exponentially
- Variance-Reduced Long-Term Rehearsal Learning with Quadratic Programming Reformulation