Calculating a most probable path

A system for determining the most probable path (MPP) comprising:

• a processor;
• a classification module to, when executed by the processor:
• determine a date and frequency of each O/D trajectory;
• classify each of the O/D trajectories into one of a plurality of periods of time based on its date and frequency;
• change a data sampling time period of any of the plurality of trajectories within the determined classification periods;
• and
• detect any O/D data descriptive of any O/D trajectory temporally longer than the determined classification periods based on context data descriptive of an aperiodic O/D trajectory;
• and
• with the classification module, calculating a route trajectory serving as a candidate of the MPP to determine a priority O/D trajectory that matches the context data, using a hidden Markov model, and
• with the classification module, providing the calculated route trajectory to a navigation system to guide a vehicle from origin to destination.

A computer program product for determining the most probable path (MPP), the computer program product comprising a non-transient computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:

• classify each of a plurality of origin and destination (O/D) trajectories into one of a plurality of periods of time;
• change a data sampling time period of any of the plurality of trajectories within the determined classification periods;
• detect any O/D data descriptive of an aperiodic O/D trajectory using context data descriptive of the aperiodic O/D trajectory, wherein an aperiodic O/D trajectory is an O/D trajectory that does not occur on a periodic basis;
• calculate a route trajectory serving as a candidate of the MPP to determine a priority O/D trajectory that matches the context data, using a deep neural network (DNN);
• and
• provide the calculated route trajectory to a navigation system to navigate a vehicle from origin to destination.

A method of calculating a most probable path (MPP) comprising:

• with an identification module, identifying a user of a computing device;
• with a classification module, classifying each of a plurality of origin and destination (O/D) trajectories into one of a plurality of periods of time based on an identity of the user, the classification comprising:
• changing a data sampling time period of any of the plurality of trajectories within the determined classification periods;
• detecting any O/D data descriptive of an aperiodic O/D trajectory using context data descriptive of the aperiodic O/D trajectory, wherein an aperiodic O/D trajectory is an O/D trajectory that occurs once during a longest classification period, wherein the context data indicates whether a date is within a day range of a holiday;
• with the classification module, detecting any O/D data descriptive of a relatively longer trajectory having a temporal longevity longer than any of the determined classification periods using context data descriptive of the relatively longer trajectory;
• with the classification module, calculating a route trajectory serving as a candidate of the MPP to determine a priority O/D trajectory that matches the context data, using a deep neural network;
• and
• navigating a vehicle from origin to destination using the MPP.