Analisis Pengaruh Kualitas Udara pada Pasien Covid-19 di Kota Berlin, Hokkaido dan Jakarta (Implementasi Model Klasifikasi: Na¨ıve Bayes, K-NN, Decision Tree dan Random Forest)

Abstract

Kualitas udara mempengaruhi kesehatan masyarakat yang tinggal di kota-kota besar terutama penyakit pernafasan dan kardiovaskular. Penelitian ini bertujuan untuk melihat hubungan antara kualitas udara terhadap tingkat keparahan pasien sindrom pernapasan akut berat (SARS-CoV-2) atau COVID-19 yang baru muncul di wilayah asia timur. Memanfaatkan algoritma klasifikasi Na¨ıve Bayes, K-nearest Neighbor, Decision Tree dan Random Forest untuk mengetahui kontribusi kualitas udara terhadap pasien covid-19. Hasil accuracy terbaik K-Nearest Neighbor dikota berlin untuk kasus dirawat sebesar 99% dengan nilai precision 79,05% dan recall 79,09%. Akurasi Decision Tree untuk kategori positif dikota hokkaido sebesar 88% dengan nilai precision 70% dan recall 71% sementara akurasi Random Forest kasus meninggal dikota jakarta 71% dengan nilai precision 57% dan recall 59%.

Air quality affects the health of society living in large cities, particularly respiratory and cardiovascular diseases. This research aims to examine the relationship between air quality and the severity of patients with severe acute respiratory syndrome (SARSCoV-2) or the emerging COVID-19 in the East Asia region. Utilizing classification algorithms such as Na¨ıve Bayes, K-nearest Neighbor, Decision Tree, and Random Forest to determine the contribution of air quality to COVID-19 patients. The best accuracy of K-Nearest Neighbor in Berlin for the hospitalized cases is 99% with a precision of 79.05% and a recall of 79.1%. The accuracy of Decision Tree for positive cases in Hokkaido is 88% with a precision of 70% and a recall of 71%, while the accuracy of Random Forest for death cases in Jakarta is 71% with a precision of 57% and a recall of 59%.