[Crib-list] CRIBB Seminar -- Speaker: Krystian Ganko (MIT) -- Friday, Feb. 2nd, 2024 @ 12:00 Noon - 1:00 PM (via ZOOM)

[Shirley Entzminger]

  VIRTUAL Seminar...


           COMPUTATIONAL RESEARCH in BOSTON and BEYOND SEMINAR
                               (CRIBB)


   ZOOM meeting info...

                    https://mit.zoom.us/j/96155042770

                       Meeting ID: 961 5504 2770

====================================

DATE:	Friday, February 2, 2024

TIME:	12:00 Noon - 1:00 PM  (Eastern Standard Time)

=====================================

SPEAKER:  Krystian Ganko  (MIT)

TITLE: System Identification of Continuous Time Compartment Models using
        the Neural Chemical Langevin Equation

ABSTRACT: 

Certain chemical and biological systems of practical interest are
experimentally observed to have both deterministic drift and diffusive
noise structure at low particle numbers, e.g., crystal nucleation and
enzymatic reaction. State-space stochastic differential equations (SDEs)
model this noise in continuous time, to more realistically describe the
stochastic dynamics compared to ordinary differential equation (ODE)
descriptions that employ the continuum approximation with no process
noise description. Moreover, for applications where process data is
bountiful, but the physics are poorly understood, neural networks may be
embedded in the state-space dynamics to help identify the system
structure (i.e., in the manner of neural state space models). However,
identifying structured noise from sampled process data in this fashion
often leads to practical identifiability issues, wherein the learned
noise models will potentially be strongly biased and uninformative.

This work explores the neural Chemical Langevin Equation (NCLE) in the
system identification of compartment models with structured colored
noise. Compared to the corresponding neural SDE (NSDE) which uses a
biased white noise model, the NCLE offers a parameterization for the
drift and diffusion structure with fewer parameters. With the help of
the Julia programming language, we construct the NCLE and NSDE variants
and deploy a particle filter-based inference approach to calibrate the
models on synthetic noisy measurement data. We then compare the
predictive performance of both models, and we finally remark on the
system identification benefits when using the NCLE model structure
variant.

----------------------------------------------------------

For a list of upcoming CRIBB Seminars, visit:

               https://math.mit.edu/crib/

=========================

  Best,

  Shirley Entzminger
  Administrative Assistant II
  MIT - Math Department
  77 Massachusetts Avenue
  Building 2, Room 350A
  Cambridge, MA 02139
  PHONE: 617-253-4994
  EMAIL: daisymae at math.mit.edu
-------------- next part --------------
A non-text attachment was scrubbed...
Name: CRIBB Seminar -- Krystian Ganko (MIT) - Friday, February 2, 2024 -.pdf
Type: application/pdf
Size: 192973 bytes
Desc: not available
URL: <http://mailman.mit.edu/pipermail/crib-list/attachments/20240131/4400b106/attachment.pdf>