A COMPUTER-IMPLEMENTED METHOD FOR TRAINING OF A MACHINE-LEARNING MODEL IN ACCIDENT WARNING LEVEL ASSESSMENT FOR A DRIVER OF A TWO-WHEELED OR A THREE-WHEELED VEHICLE

The present disclosure relates to a computer-implemented method for training of a machine-learning model in vehicle accident warning level assessment for a two-wheeled or a three-wheeled vehicle (1) that is adapted to travel in a forward movement direction (D). The method comprises preparing (S100) training data, where preparing (S100) training data comprises obtaining and automatically acquiring (S110) a plurality of sets of sensor data (100), each set of sensor data (100) being collected during a corresponding time period. The method further comprises training (S200) the machine-learning model using the prepared training data, where training (S200) the machine-learning model comprises associating (S210) each set of sensor data (100) with a certain warning level (170), among a plurality of warning levels.