arXiv research

The paper explores transferability of adversarial examples between convex and 01 loss models, finding non-transferability due to different decision boundaries caused by outliers.

Randomized Geometric Algebra for Convex Neural Networks Optimizes Transfer Learning.

Proves Riemannian positive mass theorem with singularities.

Let $U\subseteq\mathbb{R}^{n}$ be open and convex. We show that every (not necessarily Lipschitz or strongly) convex function $f:U\to\mathbb{R}$ can be approximated by real analytic convex functions, uniformly on all of $U$. In doing so we provide a technique which transfers results on uniform approximation on bounded …

End-to-end analysis of SGD for STL with adaptive sub-sampling.

FedCONST adapts update magnitudes to enhance feature generalization in FL.

When related learning tasks are naturally arranged in a hierarchy, an appealing approach for coping with scarcity of instances is that of transfer learning using a hierarchical Bayes framework. As fully Bayesian computations can be difficult and computationally demanding, it is often desirable to use posterior point es…

We study the problem of meta-learning through the lens of online convex optimization, developing a meta-algorithm bridging the gap between popular gradient-based meta-learning and classical regularization-based multi-task transfer methods. Our method is the first to simultaneously satisfy good sample efficiency guarant…

The paper proves exponential mixing for hyperbolic manifolds, with applications to geodesic holonomy.

The family of admissible positions in a transaction costs model is a random closed set, which is convex in case of proportional transaction costs. However, the convexity fails, e.g. in case of fixed transaction costs or when only a finite number of transfers are possible. The paper presents an approach to measure risks…

After defining convex near-polygons, a formula enumerating the number of triangulations of such configurations is derived in terms of edge-polynomials. The paper describes also a transfer-matrix approach for computing quantities related to triangulations.

This paper approaches the definition and properties of dynamic convex risk measures through the notion of a family of concave valuation operators satisfying certain simple and credible axioms. Exploring these in the simplest context of a finite time set and finite sample space, we find natural risk-transfer and time-co…

We establish linear regret bounds for convex smooth losses using Fenchel-Young losses.

Framework for transferring discount curve estimates across fixed-income product classes.

We propose to apply deep transfer learning from computer vision to static malware classification. In the transfer learning scheme, we borrow knowledge from natural images or objects and apply to the target domain of static malware detection. As a result, training time of deep neural networks is accelerated while high c…

Parameter-transfer is a well-known and versatile approach for meta-learning, with applications including few-shot learning, federated learning, and reinforcement learning. However, parameter-transfer algorithms often require sharing models that have been trained on the samples from specific tasks, thus leaving the task…

The paper uses thermodynamics to improve machine learning representation quality.

We propose a novel end-to-end non-minimax algorithm for training optimal transport mappings for the quadratic cost (Wasserstein-2 distance). The algorithm uses input convex neural networks and a cycle-consistency regularization to approximate Wasserstein-2 distance. In contrast to popular entropic and quadratic regular…

We consider the problem of transfer learning in an online setting. Different tasks are presented sequentially and processed by a within-task algorithm. We propose a lifelong learning strategy which refines the underlying data representation used by the within-task algorithm, thereby transferring information from one ta…

MLR-SNet learns flexible LR schedules for diverse tasks.

This paper develops convex surrogates for optimizing the multi-label F-measure.

TUSLA algorithm solves non-convex optimization problems with ReLU activations.

The localization technique from convex geometry is generalized to the setting of Riemannian manifolds whose Ricci curvature is bounded from below. In a nutshell, our method is based on the following observation: When the Ricci curvature is non-negative, log-concave measures are obtained when conditioning the Riemannian…

Study proves projective Anosov subgroups lead to mixing flows in specific spaces.

We give a new upper bound on the Selberg zeta function for a convex co-compact Schottky group acting on ${\mathbb H}^{n+1}$: in strips parallel to the imaginary axis the zeta function is bounded by $\exp (C |s|^δ)$ where $δ$ is the dimension of the limit set of the group. This bound is more precise than the optimal…

Learning-rate schedules for large models match optimization theory closely, leading to better training.

This paper studies noisy low-rank matrix completion: given partial and noisy entries of a large low-rank matrix, the goal is to estimate the underlying matrix faithfully and efficiently. Arguably one of the most popular paradigms to tackle this problem is convex relaxation, which achieves remarkable efficacy in practic…

New algorithm reduces training time for deep learning in financial hedging.

Let $Γ$ be a Zariski dense convex cocompact subgroup contained in an arithmetic lattice of $\operatorname{SO}(n, 1)^{\circ}$. We prove uniform exponential mixing of the geodesic flow for congruence covers of the hyperbolic manifold $Γ\backslash \mathbb{H}^n$ avoiding finitely many prime ideals. This extends the work of…

A new model SEQ clusters and classifies encoded features for better interpretability.

Unified theory of $θ$-expectations derived from chaotic dynamics.

Transfer learning improves portfolio optimization by identifying transfer risk.

Paper analyzes transfer risk in transfer learning for finance.

This paper explores the connection between adversarial and knowledge transferability.

Mathematical framework for transfer learning feasibility and transfer risk.

CoCoRL learns safe constraints from demonstrations with unknown rewards.

Positive weights improve kernel quadrature's accuracy.

Transfer learning borrows knowledge from a source domain to facilitate learning in a target domain. Two primary issues to be addressed in transfer learning are what and how to transfer. For a pair of domains, adopting different transfer learning algorithms results in different knowledge transferred between them. To dis…

Survey connects and systematizes transfer learning research.

Study measures impact of data and neural net similarity on transferability in restaurant sales data.

The paper analyzes phase transitions in transfer learning for perceptrons.

Adaptive source selection for positive transfer in linear models improves target dataset performance.

We consider the task of Inverse Reinforcement Learning in Contextual Markov Decision Processes (MDPs). In this setting, contexts, which define the reward and transition kernel, are sampled from a distribution. In addition, although the reward is a function of the context, it is not provided to the agent. Instead, the a…

New research on limits of transfer learning, proving key selection and dependence requirements.

This paper tackles multi-marginal optimal transport problems using DC programming.

Proposes a transfer learning method for high-dimensional quantile regression.

Localized transfer learning improves nonparametric regression performance.

Transfer learning aims at building robust prediction models by transferring knowledge gained from one problem to another. In the semantic Web, learning tasks are enhanced with semantic representations. We exploit their semantics to augment transfer learning by dealing with when to transfer with semantic measurements an…