arXiv:2607. 21773v1 Announce Type: new Abstract: In this paper, we propose and study a robust variant of the smart predict-then-optimize approach that accounts for prediction shifts due to disturbance in the covariate feature space.
arXiv:2503.00626v3 Announce Type: replace Abstract: Data-driven optimization aims to translate a machine learning model into decision-making by optimizing decisions on estimated costs. Such a pipelin...
arXiv:2609.23937v1 Announce Type: cross Abstract: Robust linear fits can resist response contamination yet remain too dense or unstable for useful global explanations. We propose penalized distillati...
arXiv:2610.07292v1 Announce Type: cross Abstract: Missing covariates are frequently encountered in supervised learning problems, and classical methods for estimation using such data use carefully cho...
arXiv:2506. 07673v2 Announce Type: replace Abstract: Large language model (LLM) evaluation is increasingly costly, prompting interest in methods that speed up evaluation by shrinking benchmark datasets.
The paper introduces Pro-Active Inference (PAI), a method for efficiently estimating a population mean when only a subset of outcomes can be observed. PAI proactively selects which observation to label next from a covariate pool, updating machine‑learning predictions online, and provides finite‑sample unbiasedness, asymptotic validity, and an improved variance estimator. Experiments on eight datasets show that PAI outperforms Sequential Active Inference in most settings, while diagnostics reveal when additional adaptivity is beneficial or detrimental.
The paper introduces Dr. Post-Training, a framework that treats general training data as a data‑induced regularizer to prevent overfitting to scarce target objectives, rather than merely a pool for selection. It constructs a feasible set of model update directions from general data and projects target‑driven updates onto this set, unifying standard training and existing data‑selection methods as special cases along a bias‑variance spectrum. Experiments on SFT, RLHF, and RLVR demonstrate that these methods outperform both standard training and state‑of‑the‑art data‑selection baselines, especially when general and target data are highly misaligned, while system optimizations keep overhead minimal.
The paper introduces Multi-Variable Conformal Prediction (MCP), a framework that jointly optimizes and calibrates multiple prediction‑set parameters in a single optimization problem, thereby eliminating the need for a calibration data split while preserving finite‑sample coverage guarantees. Two tractable variants are proposed: RemMCP, which uses constrained optimization with constraint removal and generalizes split conformal prediction, and RelMCP, which employs penalized optimization with constraint relaxation and can handle non‑convex problems. Experiments on ellipsoidal and multi‑modal prediction sets show that both RemMCP and an adaptive version of RelMCP consistently achieve target coverage with lower variance across calibration runs and, in most cases, produce smaller or comparable prediction sets compared to baseline methods.
arXiv:2606. 00563v1 Announce Type: cross Abstract: Selection bias is a common and often unavoidable aspect of real-world data that challenges the generalizability of machine learning models.
arXiv:2606. 22775v2 Announce Type: replace-cross Abstract: Distribution shift between training and deployment is a pervasive challenge for modern AI systems.
arXiv:2303. 08777v3 Announce Type: replace-cross Abstract: Cross-validation is one of the most widely used tools for risk estimation and model selection in statistics and machine learning, yet its theoretical properties when embedded in a learning procedure remain insufficiently understood.
arXiv:2607. 05806v1 Announce Type: new Abstract: Training data for machine learning is routinely collected by a selection process the model never sees: loans are observed only when granted, outcomes only when a test was ordered.