Tag
variance-reduction
- Jul 29, 2026 Split-Time Feature Subsampling in Random Forests Restricting the candidate columns at every split lowers the correlation between trees, and the strength-correlation trade explains why the default is neither one feature nor all of them. machine learning
- May 4, 2026 Variance Reduction in Stochastic Gradient Descent: SVRG and SAGA How control variates make the stochastic gradient noise vanish near the solution, why that restores a linear rate for finite sums, and what the stored state and epoch length cost. machine learning
- Apr 7, 2026 Why Monte Carlo Error Falls as the Square Root of the Sample Count The convergence rate follows from the variance of a sum, and the practical consequences are a fixed digit budget per factor of a hundred in work and an error bar that is itself estimated. scientific computing
- Jan 2, 2026 Weight Variance and Effective Sample Size in Importance Sampling Reweighting a proposal sample is exact only in the limit, and the estimator variance is governed by the second moment of the weights, which the effective sample size summarizes. scientific computing
- Aug 28, 2025 Bagging: Bootstrap Resampling and the Variance It Removes Averaging models trained on bootstrap resamples reduces variance only as far as the models are uncorrelated, which explains both the gains and the cases where there are none. machine learning
- Jul 25, 2025 Sampling BRDF Lobes and Choosing Between Them How a probability density over outgoing directions turns a reflectance integral into a weighted average, why GGX sampling has a closed form, and how lobe selection probabilities affect variance. computer graphics
- Nov 6, 2023 The Reparameterization Trick and Gradient Flow Through Sampling Why a sampling operation blocks backpropagation, how the score-function estimator works around it, and what the reparameterized estimator changes about gradient variance. deep learning
- Jun 7, 2023 Antithetic Variates and the Synchronization of Paired Runs Pairing each draw with its mirror cancels the leading fluctuation of an estimator when the integrand is monotone, and the price is a stream whose consumption must stay aligned. scientific computing