Personalized insulin dose manipulation attack and its detection using interval-based temporal patterns and machine learning algorithms

•An ML based method for accurate prediction of blood glucose (BG) level is presented.•Temporal patterns are automatically mined from raw data generated by insulin pump.•Temporal patterns are leverage by ML models to detect insulin dose manipulation attacks.•It is easier to detect overdose attack compare to underdose attack.•Different learning granularities significantly affect the detection capabilities.