-
Machine learning-based approach for online fault Diagnosis of Discrete Event System
Authors:
R Saddem,
D Baptiste
Abstract:
The problem considered in this paper is the online diagnosis of Automated Production Systems with sensors and actuators delivering discrete binary signals that can be modeled as Discrete Event Systems. Even though there are numerous diagnosis methods, none of them can meet all the criteria of implementing an efficient diagnosis system (such as an intelligent solution, an average effort, a reasonab…
▽ More
The problem considered in this paper is the online diagnosis of Automated Production Systems with sensors and actuators delivering discrete binary signals that can be modeled as Discrete Event Systems. Even though there are numerous diagnosis methods, none of them can meet all the criteria of implementing an efficient diagnosis system (such as an intelligent solution, an average effort, a reasonable cost, an online diagnosis, fewer false alarms, etc.). In addition, these techniques require either a correct, robust, and representative model of the system or relevant data or experts' knowledge that require continuous updates. In this paper, we propose a Machine Learning-based approach of a diagnostic system. It is considered as a multi-class classifier that predicts the plant state: normal or faulty and what fault that has arisen in the case of failing behavior.
△ Less
Submitted 24 October, 2022;
originally announced October 2022.
-
Malware Dynamic Analysis Evasion Techniques: A Survey
Authors:
Amir Afianian,
Salman Niksefat,
Babak Sadeghiyan,
David Baptiste
Abstract:
The Cyber world is plagued with ever-evolving malware that readily infiltrates all defense mechanisms, operates viciously unbeknownst to the user and surreptitiously exfiltrate sensitive data. Understanding the inner workings of such malware provides a leverage to effectively combat them. This understanding, is pursued through dynamic analysis which is conducted manually or automatically. Malware…
▽ More
The Cyber world is plagued with ever-evolving malware that readily infiltrates all defense mechanisms, operates viciously unbeknownst to the user and surreptitiously exfiltrate sensitive data. Understanding the inner workings of such malware provides a leverage to effectively combat them. This understanding, is pursued through dynamic analysis which is conducted manually or automatically. Malware authors accordingly, have devised and advanced evasion techniques to thwart or evade these analyses. In this paper, we present a comprehensive survey on malware dynamic analysis evasion techniques. In addition, we propose a detailed classification of these techniques and further demonstrate how their efficacy hold against different types of detection and analysis approach. Our observations attest that evasive behavior is mostly interested in detecting and evading sandboxes. The primary tactic of such malware we argue is fingerprinting followed by new trends for reverse Turing test tactic which aims at detecting human interaction. Furthermore, we will posit that the current defensive strategies beginning with reactive methods to endeavors for more transparent analysis systems are readily foiled by zero-day fingerprinting techniques or other evasion tactics such as stalling. Accordingly, we would recommend pursuit of more generic defensive strategies with emphasis on path exploration techniques that have the potential to thwart all the evasive tactics.
△ Less
Submitted 3 November, 2018;
originally announced November 2018.
-
A novel workflow for seismic net pay estimation with uncertainty
Authors:
Michael E. Glinsky,
Dale Baptiste,
Muhlis Unaldi,
Vishal Nagassar
Abstract:
This paper presents a novel workflow for seismic net pay estimation with uncertainty. It is demonstrated on the Cassra/Iris Field. The theory for the stochastic wavelet derivation (which estimates the seismic noise level along with the wavelet, time-to-depth mapping, and their uncertainties), the stochastic sparse spike inversion, and the net pay estimation (using secant areas) along with its unce…
▽ More
This paper presents a novel workflow for seismic net pay estimation with uncertainty. It is demonstrated on the Cassra/Iris Field. The theory for the stochastic wavelet derivation (which estimates the seismic noise level along with the wavelet, time-to-depth mapping, and their uncertainties), the stochastic sparse spike inversion, and the net pay estimation (using secant areas) along with its uncertainty; will be outlined. This includes benchmarking of this methodology on a synthetic model. A critical part of this process is the calibration of the secant areas. This is done in a two step process. First, a preliminary calibration is done with the stochastic reflection response modeling using rock physics relationships derived from the well logs. Second, a refinement is made to the calibration to account for the encountered net pay at the wells. Finally, a variogram structure is estimated from the extracted secant area map, then used to build in the lateral correlation to the ensemble of net pay maps while matching the well results to within the nugget of the variogram. These net pay maps are then integrated, over the area of full saturation gas, to give the GIIP distribution (Gaussian distributions for the porosity, gas expansion factor, and gas saturation for the sand end member are assumed and incorporated in the estimate of GIIP). The method is demonstrated on the Iris (UP5 turbidite) interval. The net pay is corrected for reduction in the amplitudes over part of the area due to shallow gas. The sensitivity of the GIIP to the independent stochastic variables is estimated (determining the value of information) so that business decisions can be made that maximize the value of the field.
△ Less
Submitted 1 April, 2016;
originally announced April 2016.
