Charles Webber – Detecting and Diagnosing Deterministic Language Patterns in Dynamical Systems Across Disciplines
29/09/2021, 14:00 (GMT+2, Warsaw)
Hybrid Meeting
Warsaw (Faculty of Psychology) & Online (HILL YouTube channel)
Due to covid prevention measures, there are limited places for people participating physically on the Faulty. Therefore, please fill in the registration form indicating your anticipated form of presence (online/offline): https://forms.gle/Dta5S7wLfaTmTU1b7
We are honoured to announce that the first HILL Seminar in the upcoming academic year will host Charles Webber, PhD, one of the creators of Recurrence Quantification Analysis, a propagator of the nonlinear methodology which is applicable to multiple branches of science. RQA has demonstrated broad appeal because of its mathematical simplicity in linking any dynamical system in time and space to graphical and numerical entities that are sensitive to state changes. Dr. Webber is Professor Emeritus, Cell and Molecular Physiology, Health Sciences Campus, Loyola University Chicago.
This time we will meet with you and Dr Webber in real analog world, at the Faculty of Psychology Assembly Hall (Aula), and digitally thorugh the YouTube live stream. The seminar starts at 14:00, 29.09.2021 (Wednesday). You are all cordially invited to join us. Below you will find the abstract of the lecture.
Charles L. Webber, Jr.
Loyola University Chicago, Physiology
cwebber@luc.edu
Abstract
Recurrence plots were first introduced in the context of the Hénon map (1987) and Lorentz equations (1992). But early on it became apparent that physiological processes like breathing (1994) and fatiguing muscles (1995) were amenable to this new methodology on living, dynamical systems. As the years have passed recurrence plots and their quantifications have found utility for diagnosing various systems over multiple disciplines. Why might this be the case? We posit that the popularity and success of recurrence analyses across disciplines are attributable to the common denominator of dynamical motions being present in everything. Animate and inanimate systems both wiggle in time and/or space. Parallel to the mathematical world of scalars and vectors, the paradigm of language becomes an intuitive and useful tool for explaining recurrences in terms of categorical letters and words.
In this study we demonstrate how systems speak and produce language as it were. Any detected changes in the speech reveal system states changes, some of which may be pathologic or disintegrating. Numeric data from many example systems were partitioned into 26 bins corresponding to the 26 letters in the English alphabet. Bin counts were rank-ordered and paired with decrementing letter frequencies statistically derived from almost 744 billion recurring English words (about 98 thousand distinct words) in the Google books Ngrams data base. The output stream of letters was interrupted with spaces to form (unintelligible) words according to the asymmetric (right-skewed) distribution of regular English word lengths (1 to 12 letters long). Information in these linguistic structures were studied by recurrence plots and recurrence quantifications at the orthographic and word levels (RADIUS = 0; EMBED = 1 to 20). For most systems, save random ones, shuffling of the native sequences destroyed the inherent information content of the original data (decreased DETERM).
Two pedagogical lessons stand out from this work: 1) most dynamical systems speak with meaningful contents that point to underlying deterministic rules; and 2) recurrence quantification analysis is a powerful linguistic tool irrespective of the scientific discipline employing it.
YouTube stream link: https://www.youtube.com/channel/UCyR6bJHpXQ3JfToA1S1VGeA
Phisical Venue: Assembly Hall (Aula), Faculty of Psychology, University of Warsaw
The visit is enabled by the Traincrease H2020 funding.
