Inverting convolutional neural networks for super-resolution identification of regime changes in epidemiological time series
Конференцијски прилог (Објављена верзија)
,
© 2023 Institute of Molecular Genetics and Genetic Engineering, University of Belgrade
Метаподаци
Приказ свих података о документуАпстракт
Inferring the timing and amplitude of perturbations in epidemiological systems from
their stochastically spread low-resolution outcomes is as relevant as challenging. It
is a requirement for current approaches to overcome the need to know the details
of the perturbations to proceed with the analyses. However, the general problem of
connecting epidemiological curves with the underlying incidence lacks the highly effective
methodology present in other inverse problems, such as super-resolution and dehazing
from machine vision. I will present an unsupervised physics-informed convolutional neural
network approach in reverse to connect death records with an incidence that allows the
identification of regime changes at a single-day resolution. Applied to COVID-19 data
with proper regularization and model-selection criteria, the approach can identify the
implementation and removal of lockdowns and other nonpharmaceutical interventions
with ± 0.9-day accuracy over the span of a year....
Кључне речи:
bioinformatics / physics-informed neural networks / epidemiologyИзвор:
4th Belgrade Bioinformatics Conference, 2023, 4, 13-13Издавач:
- Belgrade : Institute of molecular genetics and genetic engineering
Финансирање / пројекти:
- J.M.G.V. acknowledges support from Ministerio de Ciencia e Innovacion under grants PGC2018-101282-B-I00 and PID2021-128850NB-I00 (MCI/ AEI/FEDER, UE).
Напомена:
- Book of abstract: 4th Belgrade Bioinformatics Conference, June 19-23, 2023
Колекције
Институција/група
Institut za molekularnu genetiku i genetičko inženjerstvoTY - CONF AU - M. G. Vilar, Jose PY - 2023 UR - https://belbi.bg.ac.rs/ UR - https://imagine.imgge.bg.ac.rs/handle/123456789/1948 AB - Inferring the timing and amplitude of perturbations in epidemiological systems from their stochastically spread low-resolution outcomes is as relevant as challenging. It is a requirement for current approaches to overcome the need to know the details of the perturbations to proceed with the analyses. However, the general problem of connecting epidemiological curves with the underlying incidence lacks the highly effective methodology present in other inverse problems, such as super-resolution and dehazing from machine vision. I will present an unsupervised physics-informed convolutional neural network approach in reverse to connect death records with an incidence that allows the identification of regime changes at a single-day resolution. Applied to COVID-19 data with proper regularization and model-selection criteria, the approach can identify the implementation and removal of lockdowns and other nonpharmaceutical interventions with ± 0.9-day accuracy over the span of a year. PB - Belgrade : Institute of molecular genetics and genetic engineering C3 - 4th Belgrade Bioinformatics Conference T1 - Inverting convolutional neural networks for super-resolution identification of regime changes in epidemiological time series EP - 13 SP - 13 VL - 4 UR - https://hdl.handle.net/21.15107/rcub_imagine_1948 ER -
@conference{ author = "M. G. Vilar, Jose", year = "2023", abstract = "Inferring the timing and amplitude of perturbations in epidemiological systems from their stochastically spread low-resolution outcomes is as relevant as challenging. It is a requirement for current approaches to overcome the need to know the details of the perturbations to proceed with the analyses. However, the general problem of connecting epidemiological curves with the underlying incidence lacks the highly effective methodology present in other inverse problems, such as super-resolution and dehazing from machine vision. I will present an unsupervised physics-informed convolutional neural network approach in reverse to connect death records with an incidence that allows the identification of regime changes at a single-day resolution. Applied to COVID-19 data with proper regularization and model-selection criteria, the approach can identify the implementation and removal of lockdowns and other nonpharmaceutical interventions with ± 0.9-day accuracy over the span of a year.", publisher = "Belgrade : Institute of molecular genetics and genetic engineering", journal = "4th Belgrade Bioinformatics Conference", title = "Inverting convolutional neural networks for super-resolution identification of regime changes in epidemiological time series", pages = "13-13", volume = "4", url = "https://hdl.handle.net/21.15107/rcub_imagine_1948" }
M. G. Vilar, J.. (2023). Inverting convolutional neural networks for super-resolution identification of regime changes in epidemiological time series. in 4th Belgrade Bioinformatics Conference Belgrade : Institute of molecular genetics and genetic engineering., 4, 13-13. https://hdl.handle.net/21.15107/rcub_imagine_1948
M. G. Vilar J. Inverting convolutional neural networks for super-resolution identification of regime changes in epidemiological time series. in 4th Belgrade Bioinformatics Conference. 2023;4:13-13. https://hdl.handle.net/21.15107/rcub_imagine_1948 .
M. G. Vilar, Jose, "Inverting convolutional neural networks for super-resolution identification of regime changes in epidemiological time series" in 4th Belgrade Bioinformatics Conference, 4 (2023):13-13, https://hdl.handle.net/21.15107/rcub_imagine_1948 .