Naslov
Cancer Rates per Country - Determining the Importance of Country Level Factors using Random Forest Regressor

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

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, kratko priopćenje, znanstveni

Izvornik
First Serbian International Conference on Applied Artificial Intelligence / - Kragujevac, 1999

Skup
First Serbian International Conference on Applied Artificial Intelligence

Mjesto i datum
Kragujevac, Srbija, 19.05.2022. - 20.05.2022

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
cancer rates, correlation analysis, feature importance analysis, machine learning, random forest algorithm

Sažetak
Cancer is one of the most discussed diseases in modern healthcare. Cancer rates vary by country, this indicates that there are factors that might inuence the occurrence of cancer depending on the country. In this paper, the authors present the dataset which consists of cancer rates (CR) for 42 countries, with 10 possible country-level factors - Health Care Index (HCI), country population (POP), percentage of people living in the urban areas - urbanization rate (UR), nominal GDP (GDP), life expectance (LE), the birth rate per thousand (BR), the death rate per thousand (DR), CO2 Emission per capita (CO2), the percentage of the population that has access to the sanitation facilities (SFA), the percentage of the population that has access to a clean water source (WSA). We shall analyze the created dataset, and the influence of individual inputs is modeled and tested using the Random Forest (RF) algorithm. The results indicate that CO2 emissions, BR, and the GDP have the highest influence according to the applied RF feature importance analysis.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Računarstvo, Temeljne tehničke znanosti, Interdisciplinarne tehničke znanosti

POVEZANOST RADA