A Saddle-Point Dynamical System Approach for Robust Deep Learning [PDF]
We propose a novel discrete-time dynamical system-based framework for achieving adversarial robustness in machine learning models. Our algorithm is originated from robust optimization, which aims to find the saddle point of a min-max optimization problem
Esfandiari, Yasaman +5 more
core
Topology‐based Visual Analysis of Hydrothermal Plumes
Abstract Hydrothermal plumes are turbulent structures of intense heat and mineral smoke that rise and disperse into the deep ocean. Existing models generally characterize these systems as a single axisymmetric plume originating from a point source. However, this assumption breaks down in weakly venting, spatially distributed systems, where low‐flux ...
Adhitya Kamakshidasan +3 more
wiley +1 more source
Solution to a semilinear problem on type II regions determined by the Fucik spectrum
We prove the existence of solutions to the semi-linear problem $$displaylines{ u''(x)+alpha u^+(x)+ta u^-(x)=f(x),,quad xin (0,pi),,cr u(0)=u(pi)=0 } $$ where the point $(alpha,eta)$ falls in regions of type (II) between curves of the Fucik spectrum.
Petr Tomiczek
doaj
On parameter acceleration methods for saddle point problems
zbMATH Open Web Interface contents unavailable due to conflicting licenses.
openaire +3 more sources
Geometry‐Aware Alignment and Comparison of Hierarchical Morse Complexes with Applications
Abstract Scalar fields derived from 3D X‐ray CT scans of samples undergoing ex situ processes, such as thermal aging, chemical etching, or mechanical stress, pose unique challenges for characterizing similarities and differences across acquisitions. Typically, a sample A (source) is imaged, removed, and subjected to experimental conditions that alter ...
Aniketh Venkat +3 more
wiley +1 more source
Convergence Analysis of Stochastic Saddle Point Mirror Descent Algorithm -- A Projected Dynamical View Point [PDF]
Saddle point problems, ubiquitous in optimization, extend beyond game theory to diverse domains like power networks and reinforcement learning. This paper presents novel approaches to tackle saddle point problem, with a focus on continuous-time contexts.
Mahindrakar, Arun D +2 more
core +1 more source
Parallel Vectors Extraction using Bézier Clipping
Abstract In this paper, we propose a novel local feature extraction algorithm for the parallel vectors (PV) operator. Our method is based on Bézier clipping, which is a bracketing‐based root finding method that is commonly‐used in computer‐aided geometric design.
Nico Daßler, Tobias Günther
wiley +1 more source
Memo from Jack Imholte to Saddle Club Officers Re: Future of the Dairy Barn [PDF]
A memo from Jack Imholte to the Saddle Club officers, informing that the University still intends to demolish the Dairy Barn (the Saddle Barn), but the demolition would be happening at some indeterminate point in the future. No specific reasons are given
UMM Saddle Club
core
Dynamic Wave Trains: A Procedural Approach to Spatially Varying Ocean Synthesis
In this paper, we present a high performance procedural model capable of synthesizing and rendering a virtually infinite ocean. Our method exhibits spatial variations as the orientations, speed and choppiness of the waves is parametrized for each frequencies and each point in space. The provided visual abstract exhibits a sunset rendering of a large 40
Romain Fournier +4 more
wiley +1 more source
Monte Carlo studies of quantum cosmology by the generalized Lefschetz thimble method
Quantum cosmology aims at elucidating the beginning of our Universe. Back in early 80’s, Vilenkin and Hartle-Hawking put forward the “tunneling from nothing” and “no boundary” proposals.
Chien-Yu Chou, Jun Nishimura
doaj +1 more source

