Papers by Sayandeep Acharya
2016 Annual Conference on Information Science and Systems (CISS), 2016
IEEE journal of biomedical and health informatics, Jan 11, 2016
Humans who operate in high altitudes for prolonged durations often suffer from hypoxia. The comme... more Humans who operate in high altitudes for prolonged durations often suffer from hypoxia. The commencement of physiological and cognitive changes due to the onset of hypoxia may not be immediately apparent to the exposed individual. These changes can go unrecognized for minutes and even hours, and may lead to serious performance degradation or complete incapacitation. A dynamic system capable of monitoring and detecting decreased physiologic states due to the onset of hypoxia has the potential to prevent adverse outcomes. In this study we develop a real time hypoxia monitoring system based on a parallel M-ary decision fusion architecture. Blood oxygen saturation levels and altitude readings are the inputs and estimates of the level of hypoxia are the outputs. We develop new temporal evolution models for blood oxygen saturation and functional impairment with respect to varying altitude. The proposed models enable accurate tracking of various hypoxia levels based on the duration of stay...
Current methods for evaluating the effects of human opinions in data fusion systems are often dep... more Current methods for evaluating the effects of human opinions in data fusion systems are often dependent on human testing (which is logistically hard and difficult to arrange for repeated tests of the same population). The alternative is to use hypothetical examples, which tend to be simplistic. To facilitate studies of data fusion architectures which integrate “soft” human-generated decisions, we have used a simulator of subjective beliefs. The simulator is based on the two-stage dynamic signal detection model of Pleskac and Busemeyer (2010). We use this scheme to simulate human opinions and combine them using belief fusion methods, including Bayes' Rule; Dempster's Rule of Combination (DRC); Yager's rule; the Proportional Conflict Redistribution Rule #5 (PCR5) from Dezert-Smarandache theory; and the consensus operator from subjective logic. In our simulations, the DRC and Bayes rule exhibited performance that was on par with, and in some cases better than PCR5 and the c...
There is ongoing interest in constructing data fusion systems which are capable of using human (i... more There is ongoing interest in constructing data fusion systems which are capable of using human (i.e., soft) decisions and confidence assessments as inputs. Most relevant studies involved experimentation with humans which is often expensive, subject to strict institutional regulations, and hard to validate and replicate. Here we make use of a mathematical model of human decision-making and human confidence assessment developed by Pleskac and Busemeyer (2010) in order to compare four types of fusion operators: (1) operators that use human-subject decisions (such as the k-out-of-N majority rule); (2) operators that use subject decisions and error rates (the Chair and Varshney fusion rule); (3) operators that use subject decisions and confidence assessments (Yager's rule and the Proportional Conflict Redistribution rule #5); and (4) operators that use subject decisions, confidence assessments, and the average strength of each subject's confidence assessment, namely the average B...
2013 47th Annual Conference on Information Sciences and Systems (CISS), 2013
Current methods for evaluating the effects of human opinions in data fusion systems are often dep... more Current methods for evaluating the effects of human opinions in data fusion systems are often dependent on human testing (which is logistically hard and difficult to arrange for repeated tests of the same population). The alternative is to use hypothetical examples, which tend to be simplistic. To facilitate studies of data fusion architectures which integrate "soft" humangenerated decisions, we have used a simulator of subjective beliefs. The simulator is based on the two-stage dynamic signal detection model of Pleskac and Busemeyer (2010). We use this scheme to simulate human opinions and combine them using belief fusion methods, including Bayes' Rule; Dempster's Rule of Combination (DRC); Yager's rule; the Proportional Conflict Redistribution Rule #5 (PCR5) from Dezert-Smarandache theory; and the consensus operator from subjective logic. In our simulations, the DRC and Bayes rule exhibited performance that was on par with, and in some cases better than PCR5 and the consensus operator (when used in conjunction with a measure of source reliability). In all simulated cases, Yager's rule exhibited inferior performance.
2014 48th Annual Conference on Information Sciences and Systems (CISS), 2014
There is ongoing interest in constructing data fusion systems which are capable of using human (i... more There is ongoing interest in constructing data fusion systems which are capable of using human (i.e., soft) decisions and confidence assessments as inputs. Most relevant studies involved experimentation with humans which is often expensive, subject to strict institutional regulations, and hard to validate and replicate. Here we make use of a mathematical model of human decision-making and human confidence assessment developed by in order to compare four types of fusion operators: (1) operators that use humansubject decisions (such as the k-out-of-N majority rule); (2) operators that use subject decisions and error rates (the Chair and Varshney fusion rule); (3) operators that use subject decisions and confidence assessments (Yager's rule and the Proportional Conflict Redistribution rule #5); and (4) operators that use subject decisions, confidence assessments, and the average strength of each subject's confidence assessment, namely the average Brier scores (Dempster's rule of combination and Bayes' rule of probability combination). The ability of each fusion system to discriminate between alternatives was determined by computing the normalized area under the receiver operating characteristic curves (AUC). When the number of sources used by the fusion algorithm exceeded five, fusion operators which made use of decisions and confidence assessments alone (i.e., type (3)) produced the lowest (namely, worst) normalized AUC values. Operators which made use of subject reliabilities (i.e., types (2) and (4)) produced larger (namely, better) normalized AUC values which, in addition, were similar to those of fusion algorithms that relied on decisions alone (i.e., type (1)). For the city size discrimination task studied by Pleskac and Busmeyer, these results suggest that as the number of sources increases, the importance of decision self-assessment diminishes.
We develop models and fusion rules for oximeters that detect the onset of hypoxia. Hypoxia is a m... more We develop models and fusion rules for oximeters that detect the onset of hypoxia. Hypoxia is a medical condition affecting portions of the body that are deprived of oxygen supply. Prolonged exposure to cerebral oxygen deficiency can lead to unconsciousness or even death. The onset of hypoxia in humans is of concern for those operating in high altitudes, and in military flights characterized by high-acceleration maneuvers. Using oximeters for measuring blood oxygen saturation levels is a common means to detect hypoxia in real time. Many types of oximeters can be used for this task but all are prone to complicated noise characteristics and bias inaccuracies. It may therefore be advisable to collect and combine data streams from multiple oximeters for more reliable Hypoxia/No Hypoxia decisions (compared to decisions made by a single oximeter). Here we develop statistical noise models for three popular types of oximeters (Respironics Novametrix 515B, Nonin forehead pulse oximeter 9847, and Masimo Rad-87). We also combine data streams from these oximeters using a Kalman filter. The result is a smooth and reliable estimate of blood oxygen saturation level which can be used to detect the onset of Hypoxia.
2013 47th Annual Conference on Information Sciences and Systems (CISS), 2013
The interaction between humans and most desktop and laptop computers is often performed through t... more The interaction between humans and most desktop and laptop computers is often performed through two input devices: the keyboard and the mouse. Continuous tracking of these devices provides an opportunity to verify the identity of a user, based on a profile of behavioral biometrics from the user's previous interaction with these devices. We propose a bank of sensors, each feeding a binary detector (trying to distinguish the authentic user from all others). In this study the detectors use features derived from the keyboard and the mouse, and their decisions are fused to develop a global authentication decision. The binary classification of the individual features is developed using Naive Bayes Classifiers which play the role of local detectors in a parallel binary decision fusion architecture. The conclusion of each classifier ('authentic user' or 'other') is sent to a Decision Fusion Center (DFC) where we use the Neyman-Pearson criterion to maximize the probability of detection under an upper bound on the probability of false alarms. We compute the receiver operating characteristic (ROC) of the resulting detection scheme, and use the ROC to assess the contribution of each individual sensor to the quality of the global decision on user authenticity. In this manner we identify the characteristics (and local detectors) that are most significant to the development of correct user authentication. While the false accept rate (FAR) and false reject rate (FRR) are fixed for the local sensors, the fusion center provides trade-off between the two global error rates, and allows the designer to fix an operating point based on his/her tolerance level of false alarms. We test our approach on a real-world dataset collected from 10 office workers, who worked for a week in an office environment as we tracked their keyboard dynamics and mouse movements during interaction with laptops and desktop computers.
IT Professional, 2013
The authors apply a decision fusion architecture on a collection of behavioral biometric sensors ... more The authors apply a decision fusion architecture on a collection of behavioral biometric sensors using keystroke dynamics, mouse movement, stylometry, and Web browsing behavior. They test this active authentication approach on a dataset collected from 19 individuals in an office environment.
Computers & Electrical Engineering, 2014
Active authentication is the process of continuously verifying a user based on their ongoing inte... more Active authentication is the process of continuously verifying a user based on their ongoing interaction with a computer. In this study, we consider a representative collection of behavioral biometrics: two low-level modalities of keystroke dynamics and mouse movement, and a high-level modality of stylometry. We develop a sensor for each modality and organize the sensors as a parallel binary decision fusion architecture. We consider several applications for this authentication system, with a particular focus on secure distributed communication. We test our approach on a dataset collected from 67 users, each working individually in an office environment for a period of approximately one week. We are able to characterize the performance of the system with respect to intruder detection time and robustness to adversarial attacks, and to quantify the contribution of each modality to the overall performance.
Belief calculus provides an attractive framework to mathematically model subjective opinions of h... more Belief calculus provides an attractive framework to mathematically model subjective opinions of human observers. This work focuses on the situation when opinions provided by observers are on frames which are hierarchically related. The goal is to develop a scheme for facilitating the representation and fusion of such complex opinions in a computationally tractable way. A framework based on tree structure is proposed where, unlike in existing works in the literature, every node in itself is a frame of varied refinement and not a subset of a fixed frame. Algorithms for belief mass propagation down the tree are developed. The proposed representation is shown to be applicable to various soft-soft and hard-soft fusion situations. Using the advantages of the organization of the tree, all belief combination calculations are performed using small frames and later combined together by a simple concatenation operation making the proposed scheme a computationally attractive framework.
A group of multiple heterogeneous sensors is used to observe events of interest and their reading... more A group of multiple heterogeneous sensors is used to observe events of interest and their readings are aggregated into observation vectors that are used to draw inferences. In this generic environment we wish to integrate data provided by "hard" sensors such as readings of radar and thermal sensors with data provided by "soft" sensors such as reports from humans or context analysis by domain experts. Here we form a probabilistic representation of soft sensor data using Dempster Shafer's belief mass assignment and a consensus operator for combining human opinions with uncertainties. We then use a probability fusion rule proposed by Krzystofowicz and Long to generate a hard and soft data fusion system. This approach brings all sensor outputs to the same probabilistic framework prior to fusion. The formulation is demonstrated through three exercises involving hypothetical and real scenarios.
Aerospace Medicine and Human Performance, 2015
Uploads
Papers by Sayandeep Acharya