AI for AMR prediction

Created using ChatSlide

PerSceptoMed 1.0 offers a novel approach to addressing the global threat of Antimicrobial Resistance (AMR) by P. aeruginosa, surpassing traditional AST limitations. Developed using data from 1087 cases at IMS & SUM, it incorporates demographic specifics and features 21 antibiotics. Key preprocessing includes one-hot encoding and class imbalance solutions. The methodology employs both linear models (MLR, glmnet, SVM) and non-linear models (KNN, Decision Tree), with a 5-fold stratified cross-...