This paper presents a data-driven approach for Open Switch Faults (OSFs) classification in induction motor (IM) based EV drive. The proposed method utilizes inverter output current for classifying single and multiple OSFs, and healthy condition in IM drive. Various OSFs have been simulated under varying load and reference speed conditions. In this data-driven approach, the instantaneous currents have been utilized as feature input to Support Vector Machine (SVM), k-Nearest Neighbor (kNN), and Random Forest classifiers. Two different methods have been implemented considering single multi-class classification, and phase-wise multi-class classification approach for classifying OSFs. The performance of the SVM classifier stands out remarkable, achieving 100% classification accuracy with a single multi-class classification approach for classifying eighteen different OSFs and healthy conditions in IM drive.