Discrimination of SARS-Cov 2 and arboviruses (DENV, ZIKV and CHIKV) clinical features using machine learning techniques: a fast and inexpensive clinical screening for countries simultaneously affected by both diseases

摘要: SARS-Cov-2 (Covid-19) has spread rapidly throughout the world, and especially in tropical countries already affected by outbreaks of arboviruses, such as Dengue, Zika Chikungunya, may lead these locations to a collapse health systems. Thus, present work aims develop methodology using machine learning algorithm (Support Vector Machine) for prediction discrimination patients Covid-19 arboviruses (DENV, ZIKV CHIKV). Clinical data from 204 with both obtained 23 scientific articles 1 dataset were used. The developed model was able predict 93.1% cases 82.1% arbovirus cases, an accuracy 89.1% Area under Roc Curve 95.6%, proving be effective possible screening patients, those Covid-19, allowing early isolation.