Naslov
Spatio-temporal optimisation of SIT mosquito population control - reinforcement learning approach

Autori
Đerđ, Tamara ; Hackenberger Kutuzović, Domagoj ; Hackenberger Kutuzović, Branimir

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, neobjavljeni rad, znanstveni

Skup
International Society for Ecological Modelling Global Conference

Mjesto i datum
Toronto, Kanada, 02.05.2023. - 06.05.2023

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
reinforcement learning ; spatio-temporal optimisation ; in silico experiment ; SIT

Sažetak
The sterile insect technique (SIT) is a mosquito population control method based on the release of sterile male mosquitoes into the environment to control mosquito populations and prevent vector-borne diseases. Female mosquitoes, after mating with sterile males, produce eggs that are not viable. Therefore, this technique has the potential to be a widely used method for controlling mosquito populations. The efficiency of SIT is highly dependent on several factors, including the often highly variable environmental conditions, the mosquito population abundance and its distribution. In addition, it is crucial to estimate the right time and optimal location for the release of sterile males. Most of the mosquito population models described so far use time and some environmental data as inputs and contain fixed parameters. We have developed a dynamic, discrete, spatio-temporal matrix mosquito population model with parameters that can be automatically adjusted to new mosquito population monitoring data. By combining spatio-temporal model simulations with a reinforcement learning approach, we optimised the time, location and number of sterile males that should be released to maximise treatment efficacy in reducing mosquito population size. Based on the results of the in silico experiments conducted, we argue for adaptive rather than static (calendar-driven) mosquito management, with weather and monitoring data driving a reinforcement learning-based decision support system.

Izvorni jezik
Engleski

Znanstvena područja
Biologija, Interdisciplinarne prirodne znanosti

POVEZANOST RADA