MACSc Project Presentation - PARAMETER SENSITIVITY ANALYSIS FOR THE COMFORT LEVEL CALCULATION METHOD FOR USER BEHAVIOUR PROFILING
Who: Purvesh Patel
Title: PARAMETER SENSITIVITY ANALYSIS FOR THE COMFORT LEVEL CALCULATION METHOD FOR USER BEHAVIOUR PROFILING
Dr. Nur Zincir-Heywood - Faculty of Computer Science (Supervisor)
Dr. Raghav Sampangi - Faculty of Computer Science (Reader)
Smartphones are one of the most commonly used devices today. With the increases in the popularity of these smart devices, they are becoming obvious targets for cyber attacks. Many different research efforts are being carried out to provide more security to these devices, and one of these research efforts is comfort level based user behaviourmodeling technique for a sensor based authentication system proposed by Kayacik et al. . This method compares the current user's behaviour with the user's profile. If the user behaviour deviates from the established user pro?le, then actions such as explicit authentication will be triggered.
The motivation of my research project is to perform parameter sensitivity analysis of the comfort level based user pro?ling model proposed by Kayacik et al. . The user model method contains several parameters, each of the parameters is evaluated to study the sensitivity, and to determine its effect on the performance of the user profiling model. Then, an updated model is built incorporating the empirical findings wrt the sensitive parameters. The updated model is evaluated on different evaluation metrics such as overall comfort score, time complexity of the model, and compared against the model proposed by Kayacik et al.  using the most up to date publicly available datasets. The results of the updated model and the sensitivity of the parameters are then discussed to shed some light into the model and its usage.
Room 211, Goldberg Computer Science BuildingMACSc Project Presentation - PARAMETER SENSITIVITY ANALYSIS FOR THE COMFORT LEVEL CALCULATION METHOD FOR USER BEHAVIOUR PROFILING