检索结果-内蒙古大学图书馆

JCDL 2021 Tutorial on Systemic Challenges and Computational Solutions on Bias and Unfairness in Peer Review 21

学校读者我要写书评

暂无评论

JCDL 2021 Tutorial on Systemic Challenges and Computational ...

Proceedings of the 2021 ACM/IEEE Joint Conference on Digital Libraries

作者： Nihar B. Shah Machine Learning and Computer Science Departments Carnegie Mellon University Pittsburgh USA

ISBN: (纸本)9781665417709

Peer review is the backbone of scientific research and determines the composition of scientific digital libraries. Any systemic issues in peer review - such as biases or fraud - can systematically affect the resulting scientific digital library as well as any analyses on that library. They also affect billions of dollars in research grants made via peer review as well as entire careers of researchers. The tutorial will discuss various systemic issues in peer review via insightful experiments, several computational solutions proposed to address these issues, and a number of important open problems.A detailed writeup on the topics of this tutorial as well as a complete list of references is available in [1].

关键词：

CONFORMALIZED INTERACTIVE IMITATION learning: HANDLING EXPERT SHIFT & INTERMITTENT FEEDBACK

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhao, Michelle Simmons, Reid Admoni, Henny Ramdas, Aaditya Bajcsy, Andrea Robotics Institute School of Computer Science Carnegie Mellon University United States Departments of Statistics and Machine Learning Carnegie Mellon University United States

In interactive imitation learning (IL), uncertainty quantification offers a way for the learner (i.e. robot) to contend with distribution shifts encountered during deployment by actively seeking additional feedback from an expert (i.e. human) online. Prior works use mechanisms like ensemble disagreement or Monte Carlo dropout to quantify when black-box IL policies are uncertain;however, these approaches can lead to overconfident estimates when faced with deployment-time distribution shifts. Instead, we contend that we need uncertainty quantification algorithms that can leverage the expert human feedback received during deployment time to adapt the robot’s uncertainty online. To tackle this, we draw upon online conformal prediction, a distribution-free method for constructing prediction intervals online given a stream of ground-truth labels. Human labels, however, are intermittent in the interactive IL setting. Thus, from the conformal prediction side, we introduce a novel uncertainty quantification algorithm called intermittent quantile tracking (IQT) that leverages a probabilistic model of intermittent labels, maintains asymptotic coverage guarantees, and empirically achieves desired coverage levels. From the interactive IL side, we develop ConformalDAgger, a new approach wherein the robot uses prediction intervals calibrated by IQT as a reliable measure of deployment-time uncertainty to actively query for more expert feedback. We compare ConformalDAgger to prior uncertainty-aware DAgger methods in scenarios where the distribution shift is (and isn’t) present because of changes in the expert’s policy. We find that in simulated and hardware deployments on a 7DOF robotic manipulator, ConformalDAgger detects high uncertainty when the expert shifts and increases the number of interventions compared to baselines, allowing the robot to more quickly learn the new behavior. Project page at ***/conformalized-interactive-il/. © 2024, CC BY.

关键词： Intelligent systems

New Sequence-Independent Lifting Techniques for Cutting Planes and When They Induce Facets

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Prasad, Siddharth Vitercik, Ellen Balcan, Maria-Florina Sandholm, Tuomas Computer Science Department Carnegie Mellon University United States Management Science & Engineering and Computer Science Departments Stanford University United States Computer Science and Machine Learning Departments Carnegie Mellon University United States Computer Science Department Carnegie Mellon University Strategy Robot Inc. Strategic Machine Inc. Optimized Markets Inc. United States

Sequence-independent lifting is a procedure for strengthening valid inequalities of an integer program. We generalize the sequence-independent lifting method of Gu, Nemhauser, and Savelsbergh (GNS lifting) for cover inequalities and correct an error in their proposed generalization. We obtain a new sequence-independent lifting technique—piecewise-constant (PC) lifting—with a number of interesting properties. We derive a broad set of sufficient conditions under which PC lifting is facet defining. To our knowledge, this is the first characterization of facet-defining sequence-independent liftings that are efficiently computable from the underlying cover. Finally, we demonstrate via experiments that PC lifting can be a useful alternative to GNS lifting. We test our new lifting techniques atop a number of novel cover cut generation routines, which prove to be effective in experiments with CPLEX. © 2024, CC BY.

关键词： Integer programming

SAMPLING IS AS EASY AS learning THE SCORE: THEORY FOR DIFFUSION MODELS WITH MINIMAL DATA ASSUMPTIONS 11

学校读者我要写书评

暂无评论

SAMPLING IS AS EASY AS LEARNING THE SCORE: THEORY FOR DIFFUS...

11th International Conference on learning Representations, ICLR 2023

作者： Chen, Sitan Chewi, Sinho Li, Jerry Li, Yuanzhi Salim, Adil Zhang, Anru R. Department of EECS University of California Berkeley United States Department of Mathematics Massachusetts Institute of Technology United States Microsoft Research United States Microsoft Research Machine Learning Department Carnegie Mellon University United States Departments of Biostatistics & Bioinformatics Computer Science Mathematics and Statistical Science Duke University United States

We provide theoretical convergence guarantees for score-based generative models (SGMs) such as denoising diffusion probabilistic models (DDPMs), which constitute the backbone of large-scale real-world generative models such as DALL·E 2. Our main result is that, assuming accurate score estimates, such SGMs can efficiently sample from essentially any realistic data distribution. In contrast to prior works, our results (1) hold for an L2-accurate score estimate (rather than L∞-accurate);(2) do not require restrictive functional inequality conditions that preclude substantial non-log-concavity;(3) scale polynomially in all relevant problem parameters;and (4) match state-of-the-art complexity guarantees for discretization of the Langevin diffusion, provided that the score error is sufficiently small. We view this as strong theoretical justification for the empirical success of SGMs. We also examine SGMs based on the critically damped Langevin diffusion (CLD). Contrary to conventional wisdom, we provide evidence that the use of the CLD does not reduce the complexity of SGMs. © 2023 11th International Conference on learning Representations, ICLR 2023. All rights reserved.

关键词： Diffusion

Enhancing Peer Review in Astronomy: A machine learning and Optimization Approach to Reviewer Assignments for ALMA

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Carpenter, John M. Corvillón, Andrea Shah, Nihar B. Joint ALMA Observatory Avenida Alonso de Córdova 3107 Vitacura Santiago Chile Machine Learning and Computer Science Departments Carnegie Mellon University PittsburghPA15213 United States

The increasing volume of papers and proposals that undergo peer review emphasizes the pressing need for greater automation to effectively manage the growing scale. In this study, we present the deployment and evaluation of machine learning and optimization techniques to assign proposals to reviewers that were developed for the Atacama Large Millimeter/submillimeter Array (ALMA) during the Cycle 10 Call for Proposals issued in 2023. Using topic modeling algorithms, we identify the proposal topics and assess reviewers’ expertise based on their previous ALMA proposal submissions. We then apply an adapted version of the assignment optimization algorithm from PeerReview4All (Stelmakh et al. 2021a) to maximize the alignment between proposal topics and reviewer expertise. Our evaluation shows a significant improvement in matching reviewer expertise: the median similarity score between the proposal topic and reviewer expertise increased by 51 percentage points compared to the previous cycle, and the percentage of reviewers reporting expertise in their assigned proposals rose by 20 percentage points. Furthermore, the assignment process proved highly effective in that no proposals required reassignment due to significant mismatches, resulting in a savings of 3 to 5 days of manual effort. © 2024, CC BY.

关键词： Adversarial machine learning

Calibration with Privacy in Peer Review

学校读者我要写书评

暂无评论

Calibration with Privacy in Peer Review

IEEE International Symposium on Information Theory

作者： Wenxin Ding Gautam Kamath Weina Wang Nihar B. Shah Computer Science Department University of Chicago Computer Science Department University of Waterloo Computer Science Department Carnegie Mellon University Machine Learning and Computer Science Departments Carnegie Mellon University

ISBN: (数字)9781665421591

ISBN: (纸本)9781665421607

This paper is eligible for the Jack Keil Wolf ISIT Student Paper Award. Reviewers in peer review are often miscalibrated: they may be strict, lenient, extreme, moderate, etc. A number of algorithms have previously been proposed to calibrate reviews. Such attempts of calibration can however leak sensitive information about which reviewer reviewed which paper. In this paper, we identify this problem of calibration with privacy, and provide a foundational building block to address it. Specifically, we present a theoretical study of this problem under a simplified-yet-challenging model involving two reviewers, two papers, and an MAP-computing adversary. Our main results establish the Pareto frontier of the tradeoff between privacy (preventing the adversary from inferring reviewer identity) and utility (accepting better papers), and design explicit computationally-efficient algorithms that we prove are Pareto optimal.

关键词： Privacy Computational modeling Pareto optimization Calibration Lifting equipment Information theory

New Parameters to Monitor Rehabilitation after Tibial Fractures Based on Ground-Reaction Forces Enable Assessment of Heavily Altered Gait

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Wolff, Christian Warmerdam, Elke Dahmen, Tim Pohlemann, Tim Slusallek, Philipp Ganse, Bergita Saarbrücken Germany Saarbrücken Graduate School of Computer Science Germany Departments and Institutes of Surgery Saarland University Homburg Germany Chair for Computer Vision and Machine Learning Hochschule Aalen Aalen Germany Department of Trauma Hand and Reconstructive Surgery Departments and Institutes of Surgery Saarland University Homburg Germany

Background and Objective: Instrumented insoles have created new opportunities for patient monitoring. As the shape of the ground-reaction force curve during the stance phase is altered in patients after lower-extremity fracture, classical parameters derived from the curve often cannot be used to monitor the early rehabilitation process. We proposed new parameters applicable to individuals in the early stages of recovery after a tibial fracture and hypothesized interrelation with healing progress. Methods: In an observational prospective longitudinal study, data were collected from 13 patients with tibial fractures during straight walking at each hospital visit using instrumented insoles with 16 pressure sensors at 100 Hz. To analyse changes of the new parameters throughout healing, parametrized curves were fitted. If successful, linear, polynomial (2nd and 3rd degree), and logarithmic regression analyses were conducted to determine the best fit reflected in the highest R2-value and lowest-fitting error. A Wald Test with t-distribution was employed for statistical analysis. Strides were classified as regular or non-regular, and changes in this proportion throughout healing were analysed. Results: The parameters with the highest R2-values were the mean force between enclosing turning points (R2 = 0.715), the absolute time between enclosing turning points (R2 = 0.707), and the highest overall force (R2 = 0.722). There was a significant increase of regular strides on both the injured (R2 = 0.427, p 2 = 0.506, p © 2024, The Authors. All rights reserved.

关键词： Patient monitoring

Improved Sample Complexity Bounds for Branch-And-Cut 28

学校读者我要写书评

暂无评论

Improved Sample Complexity Bounds for Branch-And-Cut

28th International Conference on Principles and Practice of Constraint Programming, CP 2022

作者： Balcan, Maria-Florina Prasad, Siddharth Sandholm, Tuomas Vitercik, Ellen Computer Science and Machine Learning Departments Carnegie Mellon University PittsburghPA United States Computer Science Department Carnegie Mellon University PittsburghPA United States Optimized Markets Inc. PittsburghPA United States Strategic Machine Inc. PittsburghPA United States Strategy Robot Inc. PittsburghPA United States Department of Electrical Engineering and Computer Sciences University of California BerkeleyCA United States

ISBN: (纸本)9783959772402

The branch-and-cut algorithm for integer programming has a wide variety of tunable parameters that have a huge impact on its performance, but which are challenging to tune by hand. An increasingly popular approach is to use machine learning to configure these parameters based on a training set of integer programs from the application domain. We bound how large the training set should be to ensure that for any configuration, its average performance over the training set is close to its expected future performance. Our guarantees apply to parameters that control the most important aspects of branch-and-cut: node selection, branching constraint selection, and cut selection, and are sharper and more general than those from prior research. © Maria-Florina Balcan, Siddharth Prasad, Tuomas Sandholm, and Ellen Vitercik.

关键词： Integer programming

Control when confidence is costly

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Castillo, Itzel Olivos Schrater, Paul Pitkow, Xaq Department of Computer Science Rice University HoustonTX77005 United States Departments of Computer Science and Psychology University of Minnesota MinneapolisMN55455 United States Departments of Electrical and Computer Engineering and Computer Science Rice University HoustonTX77005 United States Neuroscience Institute Department of Machine Learning Carnegie Mellon University PittsburghPA15213 United States Department of Neuroscience Baylor College of Medicine HoustonTX77030 United States

We develop a version of stochastic control that accounts for computational costs of inference. Past studies identified efficient coding without control, or efficient control that neglects the cost of synthesizing information. Here we combine these concepts into a framework where agents rationally approximate inference for efficient control. Specifically, we study Linear Quadratic Gaussian (LQG) control with an added internal cost on the relative precision of the posterior probability over the world state. This creates a trade-off: an agent can obtain more utility overall by sacrificing some task performance, if doing so saves enough bits during inference. We discover that the rational strategy that solves the joint inference and control problem goes through phase transitions depending on the task demands, switching from a costly but optimal inference to a family of suboptimal inferences related by rotation transformations, each misestimate the stability of the world. In all cases, the agent moves more to think less. This work provides a foundation for a new type of rational computations that could be used by both brains and machines for efficient but computationally constrained control. © 2024, CC BY.

关键词： Stochastic systems