arXiv subjects

J Williams

Publications and source records attributed to J Williams.

4 recordsLinked to original sources

arXiv2022

The salient crossings of a crown diagram

A crown diagram of a smooth, closed oriented 4-manifold can be thought of as the projection of a link in the product of a closed surface and the circle, with chords in the circle direction connecting the strands of each crossing. This paper uses a straightforward assignment of integers to these chords to show that the smooth structures on the topological 4-manifold underlying a pair of Fintushel-Stern knot surgery 4-manifolds which are known to have the same Seiberg-Witten invariant are not isotopic. A natural question is to determine if the argument may be strengthened to show these manifolds are not diffeomorphic.

math.GT↗

arXiv2022

Depicting a generalized shift move in crown diagrams

This paper gives a diagrammatic way to perform a generalized shift move on a crown diagram of a smooth 4-manifold. Applications include a simplified proof that if two crown diagrams are related by a generalized shift move, then they are slide-equivalent; a method for converting a genus g > 1 Lefschetz fibration into a crown diagram; and the fact that the vanishing cycles of such a crown diagram are slide-equivalent to a standard inclusion of the Lefschetz vanishing cycles into a genus g + 1 surface.

math.GT↗

arXiv2017

Dynamic Mortality Risk Predictions in Pediatric Critical Care Using Recurrent Neural Networks

Viewing the trajectory of a patient as a dynamical system, a recurrent neural network was developed to learn the course of patient encounters in the Pediatric Intensive Care Unit (PICU) of a major tertiary care center. Data extracted from Electronic Medical Records (EMR) of about 12000 patients who were admitted to the PICU over a period of more than 10 years were leveraged. The RNN model ingests a sequence of measurements which include physiologic observations, laboratory results, administered drugs and interventions, and generates temporally dynamic predictions for in-ICU mortality at user-specified times. The RNN's ICU mortality predictions offer significant improvements over those from two clinically-used scores and static machine learning algorithms.

stat.ML↗

arXiv1999

Analysis of Rapidity Gap Cuts in Diffractive DIS

The requirement of a large pseudo-rapidity gap to select diffractive DIS events at HERA restricts the kinematically accessible region of phase space for a significant range of $Q^2$, $β$ and $\xpom$. Consequences of this include a breakdown of $\xpom$-factorization in large rapidity gap diffractive samples and an enhancement in the relative contribution of quark-antiquark-gluon processes over dijet processes in the diffractive DIS sample.

hep-ph↗