Spaceborne satellite images offer consistent and frequent information about surface objects, which can be utilized to explore mineral resources over a vast geographical area. To map valuable mineral resources, narrowband hyperspectral images are more efficient and accurate than broadband multispectral images. Moreover, machine learning (ML) has received remarkable attention in various fields. The spaceborne hyperspectral remote sensing and ML can improve and automate the mineral prospectivity mapping process. The present study is an attempt to map the prospectivity of hydrothermally altered and weathered minerals using a hyperspectral image captured by the recently launched PRISMA spaceborne hyperspectral sensor and various ML algorithms. The study was performed over the area of Jahazpur, Bhilwara district, India (75° 06′ 23.17′′ E, 25° 25′ 23.37′′ N). To generate the mineral distribution map for various mineral deposits, the Spectral Angle Mapper (SAM) algorithm was employed. The prepared dataset contained 173 spectral features; therefore, principal component analysis (PCA) was applied to reduce the dimension of the dataset. The generated mineral distribution map was verified through the validation survey in the study area. Various supervised classification models were developed, which were evaluated using the measures such as overall accuracy, precision, recall, F1-score, and kappa coefficient. The results revealed that the SVM model outperformed among all the classification models in terms of all the measures and it achieves overall accuracy of 99.38%. In general, the employed PRISMA hyperspectral dataset has good application prospects in the mapping of hydrothermally altered and weathered minerals using machine learning in a mountainous area.