David Uminsky
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
David Uminsky has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- James H. von Brecht
Department of Mathematics University of California Los Angeles
3 shared publications - Thomas Laurent
Research Ireland Lero & School of Computer Science and Statistics, Trinity College Dublin – Dublin, Ireland
2 shared publications - Xavier Bresson
Department of Computer Science City University of Hong Kong Hong Kong
2 shared publications - Andrea Louise Bertozzi
Department of Mathematics UCLA 90095 Los Angeles CA
1 shared publication Christopher Dunstan
University of Maryland Baltimore County Baltimore MD USA
1 shared publication- Esteban Escobar
California Polytechnic State University Pomona Pomona CA USA
1 shared publication
Latest Publications
- Open Access
The Problem with Metrics is a Fundamental Problem for AI
Optimizing a given metric is a central aspect of most current AI approaches, yet overemphasizing metrics leads to manipulation, gaming, a myopic focus on short-term goals, and other unexpected negative consequences. This poses a fundamental contradiction for AI development. Through a series of real-world case studies, we look at various aspects of where metrics go wrong in practice and aspects of how our online environment and current business practices are exacerbating these failures. Finally, we propose a framework towards mitigating the harms caused by overemphasis of metrics within AI by: (1) using a slate of metrics to get a fuller and more nuanced picture, (2) combining metrics with qualitative accounts, and (3) involving a range of stakeholders, including those who will be most impacted.
Open AccessThe Problem with Metrics is a Fundamental Problem for AI
Optimizing a given metric is a central aspect of most current AI approaches, yet overemphasizing metrics leads to manipulation, gaming, a myopic focus on short-term goals, and other unexpected negative consequences. This poses a fundamental contradiction for AI development. Through a series of real-world case studies, we look at various aspects of where metrics go wrong in practice and aspects of how our online environment and current business practices are exacerbating these failures. Finally, we propose a framework towards mitigating the harms caused by overemphasis of metrics within AI by: (1) using a slate of metrics to get a fuller and more nuanced picture, (2) combining metrics with qualitative accounts, and (3) involving a range of stakeholders, including those who will be most impacted.
- Open Access
Classification of Single-lead Electrocardiograms: TDA Informed Machine Learning
Atrial Fibrillation is a heart condition characterized by erratic heart rhythms caused by chaotic propagation of electrical impulses in the atria, leading to numerous health complications. State-of-the-art models employ complex algorithms that extract expert-informed features to improve diagnosis. In this note, we demonstrate how topological features can be used to help accurately classify single lead electrocardiograms. Via delay embeddings, we map electrocardiograms onto high-dimensional point-clouds that convert periodic signals to algebraically computable topological signatures. We derive features from persistent signatures, input them to a simple machine learning algorithm, and benchmark its performance against winning entries in the 2017 Physionet Computing in Cardiology Challenge.
Open AccessClassification of Single-lead Electrocardiograms: TDA Informed Machine Learning
Atrial Fibrillation is a heart condition characterized by erratic heart rhythms caused by chaotic propagation of electrical impulses in the atria, leading to numerous health complications. State-of-the-art models employ complex algorithms that extract expert-informed features to improve diagnosis. In this note, we demonstrate how topological features can be used to help accurately classify single lead electrocardiograms. Via delay embeddings, we map electrocardiograms onto high-dimensional point-clouds that convert periodic signals to algebraically computable topological signatures. We derive features from persistent signatures, input them to a simple machine learning algorithm, and benchmark its performance against winning entries in the 2017 Physionet Computing in Cardiology Challenge.
- 28 November 2016
- Published by IOP Publishing in Nonlinearity
- Vol. 30 (1) , 225-273
Anisotropic assembly and pattern formation
We investigate the role of anisotropy in two classes of individual-based models for self-organization, collective behavior and self-assembly. We accomplish this via first-order dynamical systems of pairwise interacting particles that incorporate anisotropic interactions. At a continuum level, these models represent the natural anisotropic variants of the well-known aggregation equation. We leverage this framework to analyze the impact of anisotropic effects upon the self-assembly of co-dimension one equilibrium structures, such as micelles and vesicles. Our analytical results reveal the regularizing effect of anisotropy, and isolate the contexts in which anisotropic effects are necessary to achieve dynamical stability of co-dimension one structures. Our results therefore place theoretical limits on when anisotropic effects can be safely neglected. We also explore whether anisotropic effects suffice to induce pattern formation in such particle systems. We conclude with brief numerical studies that highlight various aspects of the models we introduce, elucidate their phase structure and partially validate the analysis we provide.
- 28 November 2016
- Published by IOP Publishing in Nonlinearity
- Vol. 30 (1) , 225-273
Anisotropic assembly and pattern formation
We investigate the role of anisotropy in two classes of individual-based models for self-organization, collective behavior and self-assembly. We accomplish this via first-order dynamical systems of pairwise interacting particles that incorporate anisotropic interactions. At a continuum level, these models represent the natural anisotropic variants of the well-known aggregation equation. We leverage this framework to analyze the impact of anisotropic effects upon the self-assembly of co-dimension one equilibrium structures, such as micelles and vesicles. Our analytical results reveal the regularizing effect of anisotropy, and isolate the contexts in which anisotropic effects are necessary to achieve dynamical stability of co-dimension one structures. Our results therefore place theoretical limits on when anisotropic effects can be safely neglected. We also explore whether anisotropic effects suffice to induce pattern formation in such particle systems. We conclude with brief numerical studies that highlight various aspects of the models we introduce, elucidate their phase structure and partially validate the analysis we provide.
- Open Access
- 1 January 2015
- Published by International Press of Boston in Communications in Mathematical Sciences
- Vol. 13 (4) , 955-985
Ring patterns and their bifurcations in a nonlocal model of biological swarms
International Press of Boston - publishers of scholarly mathematical and scientific journals and books
- 1 January 2015
- Published by International Press of Boston in Communications in Mathematical Sciences
- Vol. 13 (4) , 955-985
Open AccessRing patterns and their bifurcations in a nonlocal model of biological swarms
International Press of Boston - publishers of scholarly mathematical and scientific journals and books
- Open Access
Multiclass Total Variation Clustering
Ideas from the image processing literature have recently motivated a new set of clustering algorithms that rely on the concept of total variation. While these algorithms perform well for bi-partitioning tasks, their recursive extensions yield unimpressive results for multiclass clustering tasks. This paper presents a general framework for multiclass total variation clustering that does not rely on recursion. The results greatly outperform previous total variation algorithms and compare well with state-of-the-art NMF approaches.
Open AccessMulticlass Total Variation Clustering
Ideas from the image processing literature have recently motivated a new set of clustering algorithms that rely on the concept of total variation. While these algorithms perform well for bi-partitioning tasks, their recursive extensions yield unimpressive results for multiclass clustering tasks. This paper presents a general framework for multiclass total variation clustering that does not rely on recursion. The results greatly outperform previous total variation algorithms and compare well with state-of-the-art NMF approaches.
- Open Access
An Adaptive Total Variation Algorithm for Computing the Balanced Cut of a Graph
We propose an adaptive version of the total variation algorithm proposed in [3] for computing the balanced cut of a graph. The algorithm from [3] used a sequence of inner total variation minimizations to guarantee descent of the balanced cut energy as well as convergence of the algorithm. In practice the total variation minimization step is never solved exactly. Instead, an accuracy parameter is specified and the total variation minimization terminates once this level of accuracy is reached. The choice of this parameter can vastly impact both the computational time of the overall algorithm as well as the accuracy of the result. Moreover, since the total variation minimization step is not solved exactly, the algorithm is not guarantied to be monotonic. In the present work we introduce a new adaptive stopping condition for the total variation minimization that guarantees monotonicity. This results in an algorithm that is actually monotonic in practice and is also significantly faster than previous, non-adaptive algorithms.
Open AccessAn Adaptive Total Variation Algorithm for Computing the Balanced Cut of a Graph
We propose an adaptive version of the total variation algorithm proposed in [3] for computing the balanced cut of a graph. The algorithm from [3] used a sequence of inner total variation minimizations to guarantee descent of the balanced cut energy as well as convergence of the algorithm. In practice the total variation minimization step is never solved exactly. Instead, an accuracy parameter is specified and the total variation minimization terminates once this level of accuracy is reached. The choice of this parameter can vastly impact both the computational time of the overall algorithm as well as the accuracy of the result. Moreover, since the total variation minimization step is not solved exactly, the algorithm is not guarantied to be monotonic. In the present work we introduce a new adaptive stopping condition for the total variation minimization that guarantees monotonicity. This results in an algorithm that is actually monotonic in practice and is also significantly faster than previous, non-adaptive algorithms.