PARAMETER OPTIMIZATION FOR EPIDEMIOLOGICAL MODEL WITH GENETIC ALGORITHM
Primary Area: optimization
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Keywords: genetic algorithm, epidemiological model, COVID-19
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
Abstract: In this study, we propose a variation of the SEIR epidemiological model, called SEPAI3R3O, and apply genetic algorithms to analyze and optimize the associated parameters. This model was developed based on the analysis of sociodemographic and behavioral data from anomalous ICDs (International Classification of Disease) and ICPCs (International Classification of Primary Care) collected from units specialized in SARS (Severe Acute Respiratory Syndrome)(i.e., specifically flu and COVID-19) in the city of Recife, located in northeast Brazil, from April $26, 2020$, to March $7, 2021$. The main objective was to understand the dynamics of disease spread and identify critical factors that influence their spread. One of these factors is the underreporting rate, estimated at around $50\%,$ which significantly increases cases due to inadequate testing. We could precisely adjust the model parameters using a genetic optimization approach, resulting in more accurate disease dynamics predictions and a more realistic view of the number of people infected by SARS. The results indicate that the SEPAI3R3O model, when optimized with genetic algorithms, could predict the spread of the disease with an effective reproduction rate $R_0$ of $3 (95\%$ CI $2.8–3.2)$ and a growth rate of $0.014 (95\%$ CI $0.013–0.015)$ for the period analyzed. With realistic data, this approach offers a valuable tool for researchers and healthcare professionals in making decisions and formulating more effective intervention strategies.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 7599
Loading