Naslov
Universal platform for mosquito population control planning using AI

Autori
Hackenberger Kutuzović, Domagoj ; Đerđ, Tamara ; 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 ; decision support system ; HPSC ; Python

Sažetak
Mosquitoes are highly effective disease vectors. Effective mosquito control planning and implementation relies heavily on the accuracy and processing of monitoring data. However, collecting high-quality data is labour-intensive and time-consuming, especially when only standard techniques are used to monitor mosquito populations. Several new technologies are enabling the development of effective tools for processing environmental and monitoring data and software platforms with high predictive power. Such predictive interactive and self-calibrating adaptive models could be valuable tools to support mosquito population management strategies. Indeed, more accurate prediction of mosquito population dynamics and spatial dispersal dynamics could help optimise the control of mosquito populations and maximise their efficiency while reducing costs In this work, we present a highly flexible and adaptable universal mosquito population control planning platform based on mosquito population simulations, IoT and satellite-based remote sensing, data mining, high-performance scientific computing, GIS and AI techniques. Our platform includes specialised CDC-based mosquito traps that use AI to collect real-time population dynamics data, as well as several software modules and subsystems. The modules and subsystems are: 1. Database ; 2. Simulation module ; 3. Data mining module ; 4. Graphical user interface and 5. Decision support system for mosquito population control based on AI. Most of the software was implemented in the Python programming language (Numpy and CuPy were used for the implementation of the population model and for high-performance scientific computing ; Flask was used for the graphical user interface, and Scikit-learn and Tensorflow/Keras were used for all machine learning and AI implementations), making the platform highly flexible for rapid changes, adaptations and upgrades.

Izvorni jezik
Engleski

Znanstvena područja
Biologija, Interdisciplinarne prirodne znanosti

POVEZANOST RADA