stat.ML
PulseAugur coverage of stat.ML — every cluster mentioning stat.ML across labs, papers, and developer communities, ranked by signal.
3 天有情绪数据
-
New measure rigorously quantifies model complexity
Researchers have developed a new, mathematically sound, and computationally efficient method for measuring model complexity. This approach, based on analyzing similarities in model gradients across different inputs, is …
-
New method simplifies causal effect estimation by relaxing assumptions
Researchers have developed a new local learning method for selecting covariates in causal effect estimation, bypassing the need for pretreatment or causal sufficiency assumptions. This approach identifies a local bounda…
-
Two gradient steps enhance feature learning in linear-width networks
This paper investigates feature learning in two-layer neural networks with a linear width, examining the impact of two gradient descent steps compared to one. The research provides a detailed spectral characterization o…
-
New algorithm precisely locates change points with bandit feedback
Researchers have developed a new adaptive algorithm for identifying multiple change points in data under bandit feedback. This algorithm aims to precisely locate discontinuities in a piecewise-constant function with min…
-
New semi-supervised kernel test integrates covariates for improved two-sample testing
Researchers have developed a new semi-supervised kernel two-sample test designed to leverage abundant unlabeled covariate data. This method aims to improve performance by incorporating covariates, which standard tests o…
-
New research characterizes mean testing limits under arbitrary truncation
This paper characterizes the fundamental limits of mean testing under arbitrary truncation, where a portion of the probability mass is hidden. The research identifies a detectability floor created by truncation bias and…
-
New statistical method enables instrumental variable analysis without structural equations
Researchers have developed a new method for instrumental variable analysis that does not require assuming the existence of exact structural equations. This approach allows for debiased inference on least-squares solutio…
-
Researchers develop new algorithm for optimal subdata selection in machine learning
Researchers have developed a new methodology for selecting optimal subsets of data when dealing with large datasets or expensive labeling. This approach, based on optimal approximate design theory, aims to retain maxima…
-
New method offers exact, invariant decomposition for network meta-analysis
Researchers have developed a new method called contrast-space projection for network meta-analysis (NMA). This technique provides an exact and invariant decomposition of direct and indirect evidence contributions within…