The Experts below are selected from a list of 114 Experts worldwide ranked by ideXlab platform
N. Asokan - One of the best experts on this subject based on the ideXlab platform.
-
Profiling Users by Modeling Web Transactions
2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS), 2017Co-Authors: Radek Tomšů, Samuel Marchal, N. AsokanAbstract:Users of electronic devices, e.g., laptop, smartphone, etc. have characteristic behaviors while surfing the Web. Profiling this behavior can help identify the person using a given device. In this paper, we introduce a technique to profile users based on their web transactions. We compute several features extracted from a sequence of web transactions and use them with one-class classification techniques to profile a user. We assess the efficacy and speed of our method at differentiating 25 synthetic users on a benchmark dataset (from a major Security Vendor) representing 6 months of web traffic monitoring from a small enterprise network.
-
ICDCS - Profiling Users by Modeling Web Transactions
2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS), 2017Co-Authors: Radek Tomsu, Samuel Marchal, N. AsokanAbstract:Users of electronic devices, e.g., laptop, smartphone, etc. have characteristic behaviors while surfing the Web. Profiling this behavior can help identify the person using a given device. In this paper, we introduce a technique to profile users based on their web transactions. We compute several features extracted from a sequence of web transactions and use them with one-class classification techniques to profile a user. We assess the efficacy and speed of our method at differentiating 25 synthetic users on a benchmark dataset (from a major Security Vendor) representing 6 months of web traffic monitoring from a small enterprise network.
Radek Tomšů - One of the best experts on this subject based on the ideXlab platform.
-
Profiling Users by Modeling Web Transactions
2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS), 2017Co-Authors: Radek Tomšů, Samuel Marchal, N. AsokanAbstract:Users of electronic devices, e.g., laptop, smartphone, etc. have characteristic behaviors while surfing the Web. Profiling this behavior can help identify the person using a given device. In this paper, we introduce a technique to profile users based on their web transactions. We compute several features extracted from a sequence of web transactions and use them with one-class classification techniques to profile a user. We assess the efficacy and speed of our method at differentiating 25 synthetic users on a benchmark dataset (from a major Security Vendor) representing 6 months of web traffic monitoring from a small enterprise network.
George Popoiu - One of the best experts on this subject based on the ideXlab platform.
-
SYNASC - Machine Learning based Malware Detection. How to Balance Memory Footprint with Model Accuracy.
2017 19th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2017Co-Authors: Dragos Teodor Gavrilut, Dan Anton, George PopoiuAbstract:Security Vendors and malware creators play a game of cat and mouse for more than two decades now. The latest area of computer science that every Security Vendor uses is machine learning. However, usage of machine learning algorithms comes with a price, especially if dealing with large amounts of data. While training such an algorithm does not have too many limitations, there are a lot of constraints regarding its practical usability (low number of false positives, good execution speed, memory footprint, etc.). This paper aims to shed some light on different optimization aspects related to machine learning models focusing on how having extra memory at hand can be used to improve the model's accuracy. We will also discuss two major implementation directions (cloud versus local) and point out advantages and disadvantages in every case.
-
Machine Learning based Malware Detection. How to Balance Memory Footprint with Model Accuracy.
2017 19th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2017Co-Authors: Dragos Teodor Gavrilut, Dan Gabriel Anton, George PopoiuAbstract:Security Vendors and malware creators play a game of cat and mouse for more than two decades now. The latest area of computer science that every Security Vendor uses is machine learning. However, usage of machine learning algorithms comes with a price, especially if dealing with large amounts of data. While training such an algorithm does not have too many limitations, there are a lot of constraints regarding its practical usability (low number of false positives, good execution speed, memory footprint, etc.). This paper aims to shed some light on different optimization aspects related to machine learning models focusing on how having extra memory at hand can be used to improve the model's accuracy. We will also discuss two major implementation directions (cloud versus local) and point out advantages and disadvantages in every case.
Samuel Marchal - One of the best experts on this subject based on the ideXlab platform.
-
Profiling Users by Modeling Web Transactions
2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS), 2017Co-Authors: Radek Tomšů, Samuel Marchal, N. AsokanAbstract:Users of electronic devices, e.g., laptop, smartphone, etc. have characteristic behaviors while surfing the Web. Profiling this behavior can help identify the person using a given device. In this paper, we introduce a technique to profile users based on their web transactions. We compute several features extracted from a sequence of web transactions and use them with one-class classification techniques to profile a user. We assess the efficacy and speed of our method at differentiating 25 synthetic users on a benchmark dataset (from a major Security Vendor) representing 6 months of web traffic monitoring from a small enterprise network.
-
ICDCS - Profiling Users by Modeling Web Transactions
2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS), 2017Co-Authors: Radek Tomsu, Samuel Marchal, N. AsokanAbstract:Users of electronic devices, e.g., laptop, smartphone, etc. have characteristic behaviors while surfing the Web. Profiling this behavior can help identify the person using a given device. In this paper, we introduce a technique to profile users based on their web transactions. We compute several features extracted from a sequence of web transactions and use them with one-class classification techniques to profile a user. We assess the efficacy and speed of our method at differentiating 25 synthetic users on a benchmark dataset (from a major Security Vendor) representing 6 months of web traffic monitoring from a small enterprise network.
Dragos Teodor Gavrilut - One of the best experts on this subject based on the ideXlab platform.
-
SYNASC - Machine Learning based Malware Detection. How to Balance Memory Footprint with Model Accuracy.
2017 19th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2017Co-Authors: Dragos Teodor Gavrilut, Dan Anton, George PopoiuAbstract:Security Vendors and malware creators play a game of cat and mouse for more than two decades now. The latest area of computer science that every Security Vendor uses is machine learning. However, usage of machine learning algorithms comes with a price, especially if dealing with large amounts of data. While training such an algorithm does not have too many limitations, there are a lot of constraints regarding its practical usability (low number of false positives, good execution speed, memory footprint, etc.). This paper aims to shed some light on different optimization aspects related to machine learning models focusing on how having extra memory at hand can be used to improve the model's accuracy. We will also discuss two major implementation directions (cloud versus local) and point out advantages and disadvantages in every case.
-
Machine Learning based Malware Detection. How to Balance Memory Footprint with Model Accuracy.
2017 19th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2017Co-Authors: Dragos Teodor Gavrilut, Dan Gabriel Anton, George PopoiuAbstract:Security Vendors and malware creators play a game of cat and mouse for more than two decades now. The latest area of computer science that every Security Vendor uses is machine learning. However, usage of machine learning algorithms comes with a price, especially if dealing with large amounts of data. While training such an algorithm does not have too many limitations, there are a lot of constraints regarding its practical usability (low number of false positives, good execution speed, memory footprint, etc.). This paper aims to shed some light on different optimization aspects related to machine learning models focusing on how having extra memory at hand can be used to improve the model's accuracy. We will also discuss two major implementation directions (cloud versus local) and point out advantages and disadvantages in every case.