FIELD: physics.
SUBSTANCE: invention relates to the computer equipment. Method of classifying a sequence of user actions comprises: obtaining a sequence of user actions, events and time, assigning default membership class sequence; obtaining information on time, user and actual membership class of sequence of actions at specified time and changing class to actual; generation of training sample containing sequences of user actions with different membership classes; splitting a sequence of user actions into sub-sequences based on predetermined parameters; determining and scoring frequency for each of classes; selection of n most informative sub-sequences indicating accessories to a given class; searching for selected most informative n subsequences in each sequence of user actions from training sample, forming a set of Boolean attributes, where each feature indicates presence of a subsequence in a sequence of actions; feeding a set of Boolean attributes as a training sample for a classification algorithm.
EFFECT: technical result is determination of features inherent to class of operations in automatic mode with high accuracy.
2 cl, 3 dwg
Title | Year | Author | Number |
---|---|---|---|
SYSTEM AND METHOD FOR DETERMINING PROCESS ASSOCIATED WITH MALWARE ENCRYPTING COMPUTER SYSTEM FILES | 2020 |
|
RU2770570C2 |
METHOD AND SYSTEM FOR STATIC ANALYSIS OF EXECUTABLE FILES BASED ON PREDICTIVE MODELS | 2020 |
|
RU2759087C1 |
METHOD AND SYSTEM FOR DETECTING EMULATED MOBILE OPERATING SYSTEM USING MACHINE LEARNING TECHNIQUES | 2018 |
|
RU2713760C1 |
METHOD FOR OBTAINING LOW-DIMENSIONAL NUMERIC REPRESENTATIONS OF SEQUENCES OF EVENTS | 2020 |
|
RU2741742C1 |
METHOD AND SYSTEM FOR BEHAVIORAL USER PROFILING BASED ON ANALYSIS OF SENSOR SIGNALS OF MOBILE DEVICE | 2023 |
|
RU2825109C1 |
METHOD OF PROTECTION AGAINST DDOS-ATTACK ON BASIS OF TRAFFIC CLASSIFICATION | 2018 |
|
RU2704741C2 |
METHOD FOR CLUSTERING SPAM EMAILS | 2021 |
|
RU2769633C1 |
METHOD AND SYSTEM FOR SUPPORTING MEDICAL DECISION MAKING USING MATHEMATICAL MODELS OF PRESENTING PATIENTS | 2017 |
|
RU2703679C2 |
METHOD AND SYSTEM FOR DETECTING MALICIOUS FILES IN A NON-ISOLATED MEDIUM | 2020 |
|
RU2722692C1 |
METHOD AND SERVER FOR DETERMINING TRAINING SET FOR MACHINE LEARNING ALGORITHM (MLA) TRAINING | 2020 |
|
RU2817726C2 |
Authors
Dates
2019-05-29—Published
2017-11-21—Filed