A method and a device for multi-class classification of an application relative to a classification of a transport mode by machine learning, including: (a) sensors configured to measure at successive instants physical quantities specific to said application, and (b) a microprocessor configured to: (b1) acquire said successive measurements from the sensors to calculate predictors from physical quantities of said measurements, (b2) construct continuous series (aj) of samples, each sample being a vector formed of a predetermined number of predictors, (b3) break down each series (aj) of samples into a set of sequences (Sk) each corresponding to a specific class among a set of predetermined classes, (b4) under-sample the sequences associated with the classes among said set of predetermined classes while conserving for each selected sequence a predetermined fraction of samples of said sequence, thereby forming classes associated with balanced samples, and (b5) construct a classifier (F) from said balanced samples by machine learning.