Detecting changes in maps of gamma spectra with Kolmogorov–Smirnov tests

Reinhart, Alex; Ventura, Valérie; Athey, Alex

doi:10.1016/j.nima.2015.09.002

Cited by 6 publications

(7 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The solid grey lines show the ROC curve when using a global (spatially invariant) estimate of the spectral density. The dotted grey lines show the ROC curves using the two-sample KS test described in Reinhart et al (2015), which also adapts spatially but does not involve any spatial smoothing. In all panels, the horizontal axis (FPR, false positive rate) is shown on a log scale, while the vertical axis (TPR, true positive rate) is shown on an ordinary scale.…”

Section: Further Details Of Bayesian Methodsmentioning

confidence: 99%

“…Therefore, to decide whether a simulatedx (s) is an anomaly, we use a one-sample Kolmogorov-Smirnov test comparing the empirical energy CDF ofx (s) to the CDF of the estimated background spectrum at site s. This is identical to the method of Chan et al (2014) with overdispersion parameter φ = 1; the generalization to other overdispersion parameters would be straightforward in principle, although we do not explore this. 2 To establish that spatial disaggregation and smoothing can improve matters, we benchmarked our method against two others:(1) The one-sample KS/Chan et al (2014) test using the global estimate of the background as the reference distribution.(2) The two-sample KS test of Reinhart et al (2015), where "sample 1" comprises the training observations from site s, and "sample 2" is the simulatedx (s) .2 Note that the authors of Chan et al (2014) simply assume that the background is known. …”

mentioning

confidence: 99%

“…(2) The two-sample KS test of Reinhart et al (2015), where "sample 1" comprises the training observations from site s, and "sample 2" is the simulatedx (s) .…”

mentioning

confidence: 99%

See 2 more Smart Citations

Multiscale Spatial Density Smoothing: An Application to Large-Scale Radiological Survey and Anomaly Detection

Tansey

Athey

Reinhart

et al. 2017

Journal of the American Statistical Association

Self Cite

View full text Add to dashboard Cite

We consider the problem of estimating a spatially varying density function, motivated by problems that arise in large-scale radiological survey and anomaly detection. In this context, the density functions to be estimated are the background gamma-ray energy spectra at sites spread across a large geographical area, such as nuclear production and waste-storage sites, military bases, medical facilities, university campuses, or the downtown of a city. Several challenges combine to make this a difficult problem. First, the spectral density at any given spatial location may have both smooth and nonsmooth features. Second, the spatial correlation in these density functions is neither stationary nor locally isotropic. Finally, at some spatial locations, there is very little data. We present a method called multiscale spatial density smoothing that successfully addresses these challenges. The method is based on recursive dyadic partition of the sample space, and therefore shares much in common with other multiscale methods, such as wavelets and Pólya-tree priors. We describe an efficient algorithm for finding a maximum a posteriori (MAP) estimate that leverages recent advances in convex optimization for non-smooth functions.We apply multiscale spatial density smoothing to real data collected on the background gamma-ray spectra at locations across a large university campus. The method exhibits state-of-the-art performance for spatial smoothing in density estimation, and it leads to substantial improvements in power when used in conjunction with existing methods for detecting the kinds of radiological anomalies that may have important consequences for public health and safety.Key words: radiological survey, density estimation, spatial statistics, Bayesian nonparametrics, total-variation denoising, fused lasso IntroductionLost or stolen radioactive sources present a challenging security problem. Widely used for industrial radiography, sterilization, and medical imaging, these sources are often poorly secured (Gaffigan, 2012) and sometimes stolen (Korshukin and Emery, 2006). To prevent dangerous accidents and to detect radiological dispersal devices (dirty bombs) before they can be used, security agencies are interested in continuously monitoring wide areas for radiation sources. A simple method is to monitor overall radiation levels. But these vary naturally in space, as different soil, stone, and building materials can contain widely different amounts of naturally occurring radioactive materials (NORM). Moreover, different detectors exhibit very different overall sensitivities to radiation. It is therefore more effective to monitor the energy spectrum of the detected radiation instead, as different radioactive materials emit gamma radiation at distinct energies.But to find a spectral anomaly, one must first know what the normal background spectrum looks like at all spatial locations. The background spectrum is the probability distribution for the energy of a gamma ray emitted by natural sources of radiation at a g...

show abstract

Section: Further Details Of Bayesian Methodsmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Multiscale Spatial Density Smoothing: An Application to Large-Scale Radiological Survey and Anomaly Detection

Tansey

Athey

Reinhart

et al. 2017

Journal of the American Statistical Association

Self Cite

View full text Add to dashboard Cite

show abstract

“…e Bernoulli process and Markov chain [34,35] were used on all network users, with mean values of [0.63, 0.48], respectively, and a threshold of 0.05, to get a better understanding of the messaging patterns and their variability. is phase monitors the activities of anomalous nodes as a group.…”

Section: Privacy Analysismentioning

confidence: 99%

Preserving Privacy in Multimedia Social Networks Using Machine Learning Anomaly Detection

Aljably

Tian

Al-Rodhaan

2020

Security and Communication Networks

View full text Add to dashboard Cite

Nowadays, user’s privacy is a critical matter in multimedia social networks. However, traditional machine learning anomaly detection techniques that rely on user’s log files and behavioral patterns are not sufficient to preserve it. Hence, the social network security should have multiple security measures to take into account additional information to protect user’s data. More precisely, access control models could complement machine learning algorithms in the process of privacy preservation. The models could use further information derived from the user’s profiles to detect anomalous users. In this paper, we implement a privacy preservation algorithm that incorporates supervised and unsupervised machine learning anomaly detection techniques with access control models. Due to the rich and fine-grained policies, our control model continuously updates the list of attributes used to classify users. It has been successfully tested on real datasets, with over 95% accuracy using Bayesian classifier, and 95.53% on receiver operating characteristic curve using deep neural networks and long short-term memory recurrent neural network classifiers. Experimental results show that this approach outperforms other detection techniques such as support vector machine, isolation forest, principal component analysis, and Kolmogorov–Smirnov test.

show abstract

“…It accepts the hypothesis that the distributions are equal if D KS is less than a threshold value D T , which is determined by choosing a significance level [17,18]. The method has been previously and successfully applied to compare gamma spectra [19]. The function determining D T for a given significance level assumes that both cumulative distribution functions are samples of the exact same parent distribution.…”

Section: Conceptmentioning

confidence: 99%

Low-resolution gamma-ray spectrometry for an information barrier based on a multi-criteria template-matching approach

Göttsche

Schirm

Glaser

2016

Nuclear Instruments and Methods in Physics Research Section A:

View full text Add to dashboard Cite

Gamma-ray spectrometry has been successfully employed to identify unique items containing special nuclear materials. Template information barriers have been developed in the past to confirm items as warheads by comparing their gamma signature to the signature of true warheads. Their development has, however, not been fully transparent, and they may not be sensitive to some relevant evasion scenarios. We develop a fully open template information barrier concept, based on low-resolution measurements, which, by design, reduces the extent of revealed sensitive information. The concept is based on three signatures of an item to be compared to a recorded template. The similarity of the spectrum is assessed by a modification of the Kolmogorov-Smirnov test to confirm the isotopic composition. The total gamma count rate must agree with the template as a measure of the projected surface of the object. In order to detect the diversion of fissile material from the interior of an item, a polyethylene mask is placed in front of the detector. Neutrons from spontaneous and induced fission events in the item produce 2.223 MeV gamma rays from neutron capture by hydrogen-1 in the mask. This peak is detected and its intensity scales with

show abstract

Detecting changes in maps of gamma spectra with Kolmogorov–Smirnov tests

Cited by 6 publications

References 16 publications

Multiscale Spatial Density Smoothing: An Application to Large-Scale Radiological Survey and Anomaly Detection

Multiscale Spatial Density Smoothing: An Application to Large-Scale Radiological Survey and Anomaly Detection

Preserving Privacy in Multimedia Social Networks Using Machine Learning Anomaly Detection

Low-resolution gamma-ray spectrometry for an information barrier based on a multi-criteria template-matching approach

Contact Info

Product

Resources

About