Search for: All records

Recent work on supervised learning [GKR+22] defined the notion of omnipredictors, i.e., predictor functions p over features that are simultaneously competitive for minimizing a family of loss functions  against a comparator class . Omniprediction requires approximating the Bayes-optimal predictor beyond the loss minimization paradigm, and has generated significant interest in the learning theory community. However, even for basic settings such as agnostically learning single-index models (SIMs), existing omnipredictor constructions require impractically-large sample complexities and runtimes, and output complex, highly-improper hypotheses. Our main contribution is a new, simple construction of omnipredictors for SIMs. We give a learner outputting an omnipredictor that is ε-competitive on any matching loss induced by a monotone, Lipschitz link function, when the comparator class is bounded linear predictors. Our algorithm requires ≈ε−4 samples and runs in nearly-linear time, and its sample complexity improves to ≈ε−2 if link functions are bi-Lipschitz. This significantly improves upon the only prior known construction, due to [HJKRR18, GHK+23], which used ≳ε−10 samples. We achieve our construction via a new, sharp analysis of the classical Isotron algorithm [KS09, KKKS11] in the challenging agnostic learning setting, of potential independent interest. Previously, Isotron was known to properly learn SIMs in the realizable setting, as well as constant-factor competitive hypotheses under the squared loss [ZWDD24]. As they are based on Isotron, our omnipredictors are multi-index models with ≈ε−2 prediction heads, bringing us closer to the tantalizing goal of proper omniprediction for general loss families and comparators. 

In the recent literature on machine learning and decision making, calibration has emerged as a desirable and widely-studied statistical property of the outputs of binary prediction models. However, the algorithmic aspects of measuring model calibration have remained relatively less well-explored. Motivated by [BGHN23], which proposed a rigorous framework for measuring distances to calibration, we initiate the algorithmic study of calibration through the lens of property testing. We define the problem of calibration testing from samples where given n draws from a distribution  on (predictions,binaryoutcomes), our goal is to distinguish between the case where  is perfectly calibrated, and the case where  is ε-far from calibration. We make the simple observation that the empirical smooth calibration linear program can be reformulated as an instance of minimum-cost flow on a highly-structured graph, and design an exact dynamic programming-based solver for it which runs in time O(nlog2(n)), and solves the calibration testing problem information-theoretically optimally in the same time. This improves upon state-of-the-art black-box linear program solvers requiring Ω(nω) time, where ω>2 is the exponent of matrix multiplication. We also develop algorithms for tolerant variants of our testing problem improving upon black-box linear program solvers, and give sample complexity lower bounds for alternative calibration measures to the one considered in this work. Finally, we present experiments showing the testing problem we define faithfully captures standard notions of calibration, and that our algorithms scale efficiently to accommodate large sample sizes. 

Prediction of surface topography in milling usually requires complex kinematics and dynamics modeling of the milling process, plus solving physical models of surface generation is a daunting task. This paper presents a multimodal data-driven machine learning (ML) method to predict milled surface topography. The proposed method predicts the height map of the surface topography by fusing process parameters and in-process acoustic information as model inputs. This method has been validated by comparing the predicted surface topography with the measured data. 

Hybrid superconductor–semiconductor materials systems are promising candidates for quantum computing applications. Their integration into superconducting electronics has enabled on-demand voltage tunability at millikelvin temperatures. Ge quantum wells have been among the semiconducting platforms interfaced with superconducting Al to realize voltage tunable Josephson junctions. Here, we explore Nb as a superconducting material in direct contact with Ge channels by focusing on the solid-state reactions at the Nb/Ge interfaces. We employ Nb evaporation at cryogenic temperatures (∼100 K) to establish a baseline structure with atomically and chemically abrupt Nb/Ge interfaces. By conducting systematic photoelectron spectroscopy and transport measurements on Nb/Ge samples across varying annealing temperatures, we elucidated the influence of Ge out-diffusion on the ultimate performance of superconducting electronics. This study underlines the need for low-temperature growth to minimize chemical intermixing and band bending at the Nb/Ge interfaces. 

Reported herein is the development of [3 + 3] cycloaddition reactions between oxyallyl cations and nitrones to yield 1,2-oxazinane heterocycles. Oxyallyl cation intermediates, generated in situ from α-tosyloxy ketones in the presence of hexafluoro-2-propanol (HFIP), a cosolvent, and a base, are found to react with a range of nitrones to afford 1,2-oxazinanes in good to high yields. The reactions are catalyzed by hydrogen-bond donors such as phenols and squaramides, and dramatically higher diastereoselectivities are observed with 4-nitrophenol.