August 2019
Intermediate to advanced
342 pages
9h 35m
English
Keystroke dynamics is based on the cadence and rhythm of keypress events that are considered unique for each student; however, these events are not directly usable by ML algorithms as they can be spoiled by a series of external random factors, such as interruptions, error corrections, or the use of keyboard special function keys such as Shift or Caps Lock.
Therefore, it is necessary to transform the corresponding raw data representing user digitations in a dataset of features that correctly represent the user's keyboard dynamics in order to clean up the data from random disturbing factors.
Read now
Unlock full access