A quantum Jensen-Shannon graph kernel using the continuous-time quantum walk

In this paper, we use the quantum Jensen-Shannon divergence as a means to establish the similarity between a pair of graphs and to develop a novel graph kernel. In quantum theory, the quantum Jensen-Shannon divergence is defined as a distance measure between quantum states. In order to compute the quantum Jensen-Shannon divergence between a pair of graphs, we first need to associate a density operator with each of them. Hence, we decide to simulate the evolution of a continuous-time quantum walk on each graph and we propose a way to associate a suitable quantum state with it. With the density operator of this quantum state to hand, the graph kernel is defined as a function of the quantum Jensen-Shannon divergence between the graph density operators. We evaluate the performance of our kernel on several standard graph datasets from bioinformatics. We use the Principle Component Analysis (PCA) on the kernel matrix to embed the graphs into a feature space for classification. The experimental results demonstrate the effectiveness of the proposed approach. © 2013 Springer-Verlag.

A continuous-time quantum walk kernel for unattributed graphs

Kernel methods provide a way to apply a wide range of learning techniques to complex and structured data by shifting the representational problem from one of finding an embedding of the data to that of defining a positive semidefinite kernel. In this paper, we propose a novel kernel on unattributed graphs where the structure is characterized through the evolution of a continuous-time quantum walk. More precisely, given a pair of graphs, we create a derived structure whose degree of symmetry is maximum when the original graphs are isomorphic. With this new graph to hand, we compute the density operators of the quantum systems representing the evolutions of two suitably defined quantum walks. Finally, we define the kernel between the two original graphs as the quantum Jensen-Shannon divergence between these two density operators. The experimental evaluation shows the effectiveness of the proposed approach. © 2013 Springer-Verlag.

Manifold learning and the quantum Jensen-Shannon divergence kernel

The quantum Jensen-Shannon divergence kernel [1] was recently introduced in the context of unattributed graphs where it was shown to outperform several commonly used alternatives. In this paper, we study the separability properties of this kernel and we propose a way to compute a low-dimensional kernel embedding where the separation of the different classes is enhanced. The idea stems from the observation that the multidimensional scaling embeddings on this kernel show a strong horseshoe shape distribution, a pattern which is known to arise when long range distances are not estimated accurately. Here we propose to use Isomap to embed the graphs using only local distance information onto a new vectorial space with a higher class separability. The experimental evaluation shows the effectiveness of the proposed approach. © 2013 Springer-Verlag.

On the Range and the Kernel of Derivations

2000 Mathematics Subject Classification: Primary 47B47, 47B10; Secondary 47A30.

Large and Moderate Deviations Principles for Recursive Kernel Estimator of a Multivariate Density and its Partial Derivatives

2000 Mathematics Subject Classification: 62G07, 60F10.

The kernel is in the least core for permutation games

Permutation games are totally balanced transferable utility cooperative games arising from certain sequencing and re-assignment optimization problems. It is known that for permutation games the bargaining set and the core coincide, consequently, the kernel is a subset of the core. We prove that for permutation games the kernel is contained in the least core, even if the latter is a lower dimensional subset of the core. By means of a 5-player permutation game we demonstrate that, in sense of the lexicographic center procedure leading to the nucleolus, this inclusion result can not be strengthened. Our 5-player permutation game is also an example (of minimum size) for a game with a non-convex kernel.

Integrity-based kernel malware detection

Kernel-level malware is one of the most dangerous threats to the security of users on the Internet, so there is an urgent need for its detection. The most popular detection approach is misuse-based detection. However, it cannot catch up with today's advanced malware that increasingly apply polymorphism and obfuscation. In this thesis, we present our integrity-based detection for kernel-level malware, which does not rely on the specific features of malware. ^ We have developed an integrity analysis system that can derive and monitor integrity properties for commodity operating systems kernels. In our system, we focus on two classes of integrity properties: data invariants and integrity of Kernel Queue (KQ) requests. ^ We adopt static analysis for data invariant detection and overcome several technical challenges: field-sensitivity, array-sensitivity, and pointer analysis. We identify data invariants that are critical to system runtime integrity from Linux kernel 2.4.32 and Windows Research Kernel (WRK) with very low false positive rate and very low false negative rate. We then develop an Invariant Monitor to guard these data invariants against real-world malware. In our experiment, we are able to use Invariant Monitor to detect ten real-world Linux rootkits and nine real-world Windows malware and one synthetic Windows malware. ^ We leverage static and dynamic analysis of kernel and device drivers to learn the legitimate KQ requests. Based on the learned KQ requests, we build KQguard to protect KQs. At runtime, KQguard rejects all the unknown KQ requests that cannot be validated. We apply KQguard on WRK and Linux kernel, and extensive experimental evaluation shows that KQguard is efficient (up to 5.6% overhead) and effective (capable of achieving zero false positives against representative benign workloads after appropriate training and very low false negatives against 125 real-world malware and nine synthetic attacks). ^ In our system, Invariant Monitor and KQguard cooperate together to protect data invariants and KQs in the target kernel. By monitoring these integrity properties, we can detect malware by its violation of these integrity properties during execution.^

Optimal kernel development for real-time communications

The purpose of this research is to develop an optimal kernel which would be used in a real-time engineering and communications system. Since the application is a real-time system, relevant real-time issues are studied in conjunction with kernel related issues. The emphasis of the research is the development of a kernel which would not only adhere to the criteria of a real-time environment, namely determinism and performance, but also provide the flexibility and portability associated with non-real-time environments. The essence of the research is to study how the features found in non-real-time systems could be applied to the real-time system in order to generate an optimal kernel which would provide flexibility and architecture independence while maintaining the performance needed by most of the engineering applications. Traditionally, development of real-time kernels has been done using assembly language. By utilizing the powerful constructs of the C language, a real-time kernel was developed which addressed the goals of flexibility and portability while still meeting the real-time criteria. The implementation of the kernel is carried out using the powerful 68010/20/30/40 microprocessor based systems.

(Table S1) Brominated flame retardant concentrations in an ice core from Holtedahlfonna, Svalbard

Brominated flame retardants (BFRs) have been found in Arctic wildlife, lake sediment, and air. To identify the atmospheric BFR deposition history on Svalbard, Norway, we analyzed 19 BFRs, including hexabromocyclododecane (HBCD), 1,2-bis(2,4,6-tribromophenoxy)ethane (BTBPE), decabromodiphenyl ethane (DBDPE), pentabromoethylbenzene (PBEB),and 15 polybrominated diphenyl ether congeners (PBDE) in the upper 34 m of an ice core (representing 1953-2005) from Holtedahlfonna, the western-most ice sheet on Svalbard. All of the non-PBDE compounds were detected in nearly continuous profiles in the core. Seven PBDEs were not observed above background (28,47,66,100,99,154,153), while 4 were found in 1 or 2 of 6 segments (17,85,138,183). BDEs-49,71,190,209 had nearly continuous profiles but only BDE-209 in large amounts. The greatest inputs were HBCD and BDE-209, 910, and 320 pg/cm**2/yr from 1995-2005. DBDPE, BTBPE, and PBEB show nearly continuous input growth in recent core segments, but all were <6 pg/cm**2/yr. Long-range atmospheric processes may have moved these particle-bound BFRs to the site, probably during the Arctic haze season. Average air mass trajectories over 10 years show >75% of atmospheric flow to Holtedahlfonna coming from Eurasia during haze periods (March and April).

The Effects of Brominated Flame Retardants on Thyroid Hormone Homeostasis in Human Placenta Tissues and Cell Culture

Polybrominated diphenyl ethers (PBDEs) are a class of brominated flame retardants (BFRs) that have been heavily used in consumer products such as furniture foams, plastics, and textiles since the mid-1970’s. BFRs are added to products in order to meet state flammability standards intended to increase indoor safety in the event of a fire. The three commercial PBDE mixtures, Penta-, Octa-, and DecaBDE, have all been banned in the United States, however, limited use of DecaBDE is still permitted. PBDEs were phased out of production and added to the Stockholm Convention due to concerns over their environmental persistence and toxicity. Human exposure to PBDEs occurs primarily through the inadvertent ingestion of contaminated house dust, as well as though dietary sources. Despite the phase-out and discontinued use of PBDEs, human exposure to this class of chemicals is likely to continue for decades due to the continued use of treated products and existing environmental reservoirs of PBDEs. Extensive research over the years has shown that PBDEs disrupt thyroid hormone (TH) levels and neurodevelopmental endpoints in rodent and fish models. Additionally, there is growing epidemiological evidence linking PBDE exposure in humans to altered TH homeostasis and neurodevelopmental impairments in children. Due to the importance of THs throughout gestation, there is a great need to understand the effects of BFRs on the developing fetus. Specifically, the placenta plays a critical role in the transport, metabolism, and delivery of THs to the fetal compartment during pregnancy and is a likely target for BFR bioaccumulation and endocrine disruption. The central hypothesis of this dissertation research is that BFRs disrupt the activity of TH sulfotransferase (SULT) enzymes, thereby altering TH concentrations in the placenta.

In the first aim of this dissertation research, the concentrations of PBDEs and 2,4,6-TBP were measured in a cohort of 102 placenta tissue samples from an ongoing pregnancy cohort in Durham, NC. Methods were developed for the extraction and analysis of the BFR analytes. It was found that 2,4,6-TBP was significantly correlated with all PBDE analytes, indicating that 2,4,6-TBP may share common product applications with PBDEs or that 2,4,6-TBP is a metabolite of PBDE compounds. Additionally, this was the first study to measure 2,4,6-TBP in human placenta tissues.

In the second aim of this dissertation research, the placenta tissue concentrations of THs, as well as the endogenous activity of deiodinase (DI) and TH SULT enzymes were quantified using the same cohort of 102 placenta tissue samples. Enzyme activity was detected in all samples and this was the first study to measure TH DI and SULT activity in human placenta tissues. Enzyme activities and TH concentrations were compared with BFR concentrations measured in Aim 1. There were few statistically significant associations observed for the combined data, however, upon stratifying the data set based on infant sex, additional significant associations were observed. For example, among males, those with the highest concentrations of BDE-99 in placenta had T3 levels 0.80 times those with the lowest concentration of BDE-99 (95% confidence interval (CI): 0.59, 1.07). Whereas females with the highest concentrations of BDE-99 in placenta had T3 levels 1.50 times those with the lowest concentration of BDE-99 (95% CI: 1.10, 2.04). Additionally, all BFR analyte concentrations were higher in the placenta of males versus females and they were significantly higher for 2,4,6-TBP and BDE-209. 3,3’-T2 SULT activity was significantly higher in female placenta tissues, while type 3 DI activity was significantly higher in male placenta tissues. This research is the first to show sex-specific differences in the bioaccumulation of BFRs in human placenta tissue, as well as differences in TH concentrations and endogenous DI and SULT activity. The underlying mechanisms of these observed sex differences warrant further investigation.

In the third aim of this dissertation research, the effects of BFRs were examined in a human choriocarcinoma placenta cell line, BeWo. Michaelis-Menten parameters and inhibition curves were calculated for 2,4,6-TBP, 3-OH BDE-47, and 6-OH BDE-47. 2,4,6-TBP was shown to be the most potent inhibitor of 3,3’-T2 SULT activity with a calculated IC50 value of 11.6 nM. It was also shown that 2,4,6-TBP and 3-OH BDE-47 exhibit mixed inhibition of 3,3’-T2 sulfation in BeWo cell homogenates. Next, a series of cell culture exposure experiments were performed using 1, 6, 12, and 24 hour exposure durations. Once again, 2,4,6-TBP was shown to be the most potent inhibitor of basal 3,3’-T2 SULT activity by significantly decreasing activity at the high and medium dose (1 M and 0.5 M, respectively) at all measured time points. Interestingly, BDE-99 was also shown to inhibit basal 3,3’-T2 SULT activity in BeWo cells following the 24 hour exposure, despite exhibiting no inhibitory effects in the BeWo cell homogenate experiments. This indicates that BDE-99 must act through a pathway other than direct enzyme inhibition. Following exposures, the TH concentrations in the cell culture growth media and mRNA expression of TH-related genes were also examined. There was no observed effect of BFR treatment on these endpoints. Future work should focus on determining the downstream biological effects of TH SULT disruption in placental cells, as well as the underlying mechanisms of action responsible for reductions in basal TH SULT activity following BFR exposure.

This was one of the first studies to measure BFRs in a cohort of placenta tissue samples from the United States and the first study to measure THs, DI activity, and SULT activity in human placenta tissues. This research provides a novel contribution to our growing understanding of the effects of BFRs on TH homeostasis within the human placenta, and provides further evidence for sex-specific differences within this important organ. Future research should continue to investigate the effects of environmental contaminants on TH homeostasis within the placenta, as this represents the most critical and vulnerable stage of human development.

Integrity-Based Kernel Malware Detection

Kernel-level malware is one of the most dangerous threats to the security of users on the Internet, so there is an urgent need for its detection. The most popular detection approach is misuse-based detection. However, it cannot catch up with today's advanced malware that increasingly apply polymorphism and obfuscation. In this thesis, we present our integrity-based detection for kernel-level malware, which does not rely on the specific features of malware. We have developed an integrity analysis system that can derive and monitor integrity properties for commodity operating systems kernels. In our system, we focus on two classes of integrity properties: data invariants and integrity of Kernel Queue (KQ) requests. We adopt static analysis for data invariant detection and overcome several technical challenges: field-sensitivity, array-sensitivity, and pointer analysis. We identify data invariants that are critical to system runtime integrity from Linux kernel 2.4.32 and Windows Research Kernel (WRK) with very low false positive rate and very low false negative rate. We then develop an Invariant Monitor to guard these data invariants against real-world malware. In our experiment, we are able to use Invariant Monitor to detect ten real-world Linux rootkits and nine real-world Windows malware and one synthetic Windows malware. We leverage static and dynamic analysis of kernel and device drivers to learn the legitimate KQ requests. Based on the learned KQ requests, we build KQguard to protect KQs. At runtime, KQguard rejects all the unknown KQ requests that cannot be validated. We apply KQguard on WRK and Linux kernel, and extensive experimental evaluation shows that KQguard is efficient (up to 5.6% overhead) and effective (capable of achieving zero false positives against representative benign workloads after appropriate training and very low false negatives against 125 real-world malware and nine synthetic attacks). In our system, Invariant Monitor and KQguard cooperate together to protect data invariants and KQs in the target kernel. By monitoring these integrity properties, we can detect malware by its violation of these integrity properties during execution.

Brominated flame retardants and biomagnification factors in marine species in the North-East Atlantic

BACKGROUND: Concentrations of brominated flame retardants (BFRs) are reported to increase in marine ecosystems. OBJECTIVES: Characterize exposure to BFRs in animals from different trophic levels in North-East Atlantic coastal marine ecosystems along a latitudinal gradient from southern Norway to Spitsbergen, Svalbard, in the Arctic. Calanoid species were collected from the Oslofjord (59°N), Froan (64°N), and Spitsbergen (> 78°N); Atlantic cod (Gadus morhua) from the Oslofjord and Froan; polar cod (Boreogadus saida) from Bear Island (74°N) and Spitsbergen; harbor seal (Phoca vitulina) from the Oslofjord, Froan, and Spitsbergen; and ringed seal (Phoca vitulina) from Spitsbergen. Eggs of common tern (Sterna hirundo) were collected from the Oslofjord, and eggs of arctic terns (Sterna paradisaea) from Froan and Spitsbergen. RESULTS: Levels of polybrominated diphenylethers (PBDEs) and hexabromocyclododecane (HBCD) generally decreased as a function of increasing latitude, reflecting distance from release sources. The clear latitudinal decrease in levels of BFRs was not pronounced in the two tern species, most likely because they are exposed during migration. The decabrominated compound BDE-209 was detected in animals from all three ecosystems, and the highest levels were found in arctic tern eggs from Spitsbergen. HBCD was found in animals from all trophic levels, except for in calanoids at Froan and Spitsbergen. CONCLUSIONS: Even though the levels of PBDEs and HBCD are generally low in North-East Atlantic coastal marine ecosystems, there are concerns about the relatively high presence of BDE-209 and HBCD.

Flammability Standards – A Major Role in Flame Retardant Exposure? Televisions as a case study.

Flame retardants (FRs) are added to materials to enhance the fire safety level of readily combustible polymers. Although they have been purported to aid in preventing fires in some cases, they have also become a significant cause for concern given the vast data on environmental persistence and human and animal adverse health effects. Evidence since the 1980s has shown that Canadian, American and Europeans have detectable levels of FRs in their bodies. North Americans in particular have high levels of these chemicals due to stringent flammability standards and the higher use of polybrominated diphenyl ethers (PBDEs) in North America as opposed to Europe. FRs have been detected in household dust and some evidence suggests that TVs could be a significant source of exposure to FRs. It is imperative to re-visit the flammability standard (UL94V) that allows for FR use in TVs plastic materials by providing a risk versus benefit analysis to determine if this standard provides a fire safety benefit and if it plays a major role in FR exposure. This report first examined the history of televisions and the progression to the UL94V flammability test standard to understand why FRs were first added to polymers used in the manufacturing of TVs. It has been demonstrated to be due to fire hazards resulting from the use of plastic materials in cathode-ray tube (CRT) TVs that had an “instant-on” feature and high voltage and operating temperatures. In providing a risk versus benefit analysis, this paper presents the argument that 1) by providing a market survey the current flammability test standard (UL94V) is outdated and lacks relevance to current technology as flat, thin, energy efficient Liquid Crystal Displays (LCDs) dominate over traditionally used heavy, bulky and energy-intensive CRTs; 2) FRs do not impart fire safety benefits considering that there is a lack of valid fire safety concern, such as reduced internal and external ignition and fire hazard, and a lack of valid fire data and hazard for television fires in general and finally; 3) the standard is overly stringent as it does not consider the risk due to exposure to FRs in household dust due to the proliferation and greater use of televisions in households. Therefore, this report argues that the UL94V standard has become trapped in history and needs to be updated as it may play a major role in FR exposure.

Evaluation of the Genotoxic and Physiological Effects of Decabromodiphenyl Ether (BDE-209) and Dechlorane Plus (DP) Flame Retardants in Marine Mussels (Mytilus galloprovincialis)

Dechlorane Plus (DP) is a proposed alternative to the legacy flame retardant decabromodiphenyl ether (BDE-209), a major component of Deca-BDE formulations. In contrast to BDE-209, toxicity data for DP are scarce and often focused on mice. Validated dietary in vivo exposure of the marine bivalve (Mytilus galloprovincialis) to both flame retardants did not induce effects at the physiological level (algal clearance rate), but induced DNA damage, as determined by the comet assay, at all concentrations tested. Micronuclei formation was induced by both DP and BDE-209 at the highest exposure concentrations (100 and 200 mu g/L, respectively, at 18% above controls). DP caused effects similar to those by BDE-209 but at lower exposure concentrations (5.6, 56, and 100 mu g/L for DP and 56, 100, and 200 mu g/L for BDE-209). Moreover, bioaccumulation of DP was shown to be concentration dependent, in contrast to BDE-209. The results described suggest that DP poses a greater genotoxic potential than BDE-209