Naslov
Utilization of Genetic Programming for Estimation of Molecular Structures Ground State Energies

Autori
Anđelić, Nikola ; Baressi Šegota, Sandi ; Lorencin Ivan ; Glučina, Matko ; Musulin, Jelena ; Štifanić, Daniel ; Car, Zlatan

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
1st Serbian International Conference on Applied Artificial Intelligence (SICAAI) / - , 2022, 1-4

Skup
1st Serbian International Conference on Applied Artificial Intelligence (SICAAI)

Mjesto i datum
Kragujevac, Srbija, 19.05.2022. - 20.05.2022

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
CHNOPS dataset, Genetic Programming, Ground State Energies

Sažetak
In this paper, we shall be utilizing genetic programming (GP) to predict ground-state energies of molecules made up of C, H, N, O, P, and S (CHONPS) atoms. The GP was trained and tested on a publicly available dataset which consist of 16242 molecules where ground state energies were computed using the density functional theory (DFT). The optimal parameters of GP were chosen using the random parameter search method. After multiple GP executions, the best symbolic expression was chosen using a coefficient of determination (R2), mean absolute error (MAE), and root mean square error (RMSE). The best symbolic expression achieved R2, MAE, and RMSE of 0.9434, 0.48, and 0.86, respectively.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Računarstvo, Temeljne tehničke znanosti

POVEZANOST RADA