JIITA, Vol.10 no.1 pp.1250-1262 (2026), DOI: 10.22664/ISITA.2026.10.1.1250
G. Ezhilvani, J. Wessly, and R. Durga
Abstract. Nerve affliction is individual of ultimate famous and poisonous ailments in the concern, and in abundance arrangement diminish their existences from that affliction every twelvemonth. Premature understanding concerning this affliction is aware drop family’s lives. ML, a stylized discourse demesne, is human of net seasonable, fastest, and low-cost explanation to observation affliction. In this opine almost, we aim to get an ML influence that can meet congestive nerve nonstarter incidental the highest available kill utilizing the Chairman insight ill Dataset. This countenance in those datasets resorted to take the pretend and the belief. To refrain overfitting (on account of the exclamation of dimensionality) on account of the Brobdingnagian decide of visage in the President dataset, these datasets existed shortened to a depreciate spatial subspace using the Siphonophore bettering treasure. The Medusan treasure has a superior touching percentage and is versatile to mature the importance looks. The models acquired by nurture the feature-picked dataset with opposition ML algorithms were demonstrated, and their acts were dignified. The maximum appear was got for the SVM classifier Modify of 98.56%, 98.37%, 98.47%, and 94.48%, respectively. This outcome dissembling that the connexion of the Medusa amendatory innovation and SVM classifier has the maximum calculation for use in suspicion affliction rendition.
Keywords; Machine Learning; Dataset; Medical Diagnosis; Feature Selection; Dimensionality reduction; Support Vector Machine
Fullpaper: