System and method for detecting anomaly in a handheld device

A system for detecting anomaly in a hand-held device comprising:

• a memory storing observations of one or more parameters with respect to time, wherein said parameters are based on one or more of user's mobility pattern, device usage context, and device behaviour;
• an anomaly detection engine coupled with the memory and configured to select and extract a first set of parameters, wherein said anomaly detection engine comprises:
• an index unit configured to construct one or more indexes based on said first set of parameters;
• a time series model unit configured to fit a time-series on each constructed index;
• an outlier detection unit configured to detect outlier observations for each index based on said fitted time series model;
• a dummy variable assigner configured to associate each observation with a dummy variable value based on detected outlier observations;
• and
• a calculation unit to detect an anomaly based on said dummy variable values and said set of parameters.