Attention-based Graph ResNet with focal loss for epileptic seizure detection

Epilepsy is a chronic brain disease resulted from the central nervous system lesion, which leads to repeated seizure occurs for the patients. Automatic seizure detection with Electroencephalogram (EEG) has witnessed great progress. However, existing methods paid little attention to the topological relationships of different EEG electrodes. Latest neuroscience researches have demonstrated the connectivity between different brain regions. Besides, class-imbalance is a common problem in EEG based seizure detection. The duration of epileptic EEG signals is much shorter than that of normal signals. In order to deal with the above mentioned two challenges, we propose to model the multi-channel EEG data using the Attention-based Graph ResNet (AGRN). In particular, each channel of the EEG signal represents a node of the graph and the inter-channel relations are modeled via the adjacency matrix in the graph. The loss function of the ARGN model is re-designed using focal loss to cope with the class-imbalance problem. The proposed ARGN with focal model could learn discriminative features from the raw EEG data. Experiments are carried out on the CHB-MIT dataset. The proposed model achieves an average accuracy of 98.70%, a sensitivity of 97.94%, a specificity of 98.66% and a precision of 98.62%. The Area Under the ROC Curve (AUC) is 98.69%.

ESTABLISHING INDOOR SUBSPACING REQUIREMENTS OF AN LOD (LEVEL OF DETAIL) MODEL FOR GENERATING NETWORK-BASED TOPOLOGICAL DATA

Nowadays, the complexity of structures in urban environments and the interest in location-based applications increase simultaneously. Along with this is the rise in demand for the firm establishment of data models representing these spaces. Establishing network models that portray topological relationships of space have strengthened support for navigation applications. However, researchers have revisited the limitations of existing standards. As analogous standards have specifications for expressing space at various scales, most have focused on outdoor space or the geometric aspect. Hence, this paper proposes subspacing requirements for a Level of Detail (LOD) model for network-based topological data. We examine various constraints that influence space partition and align these with various application cases for indoor navigation. Through these, we investigate appropriate space subdivision approaches for each level according to applicable constraints and recommended applications. This study poses as an initial study towards establishing a general framework for implementing a 3D hierarchical network-based topological data model.

Zum variationslinguistischen Verhältnis von Stadt und Land.

In this paper, we explore the geolinguistic relationship between urban and rural areas through the conceptualisation and modeling of spatial topologies. Geolinguistic topologies concern the structure of the mutual linguistic relationship between localities. They can be defined either deductively or on the basis of empirical data and represent the linguistic similarities or distances between localities. We operationalise and apply several such topological models to Austrian data from the Atlas zur deutschen Alltagssprache (AdA), a linguistic atlas documenting colloquial German using crowd-sourcing methods. The results are evaluated on the basis of statistical examination and of visualisations of the topological relationships predicted by the models. It is confirmed that linguistic similarity is determined both by geographical distance and by the distribution of population, but the exact relationship is complex: Not only do smaller geographic distances on the one hand and higher population numbers on the other hand bring about increased linguistic similarity; the relevance of these two factors for linguistic similarity varies with population size, too, such that linguistic relationships between cities are determined more by their size and less by their distance, while for smaller locations the opposite is true. Hence, no single topological model can be identified as superior; instead, the individual models emphasise different aspects of the linguistic relationship between urban and regional language usage.

Analysis of Carrier-Based Aircraft Catapult Launching Based on Variable Topology Dynamics

The catapult process of a carrier-based aircraft includes multiple links such as catapult tensioning, separation of the holding rod, dragging and running, separation of the catapult and drag shuttle, and free running. The connection relationships between the front landing gear of the carrier-based aircraft and other related components in each link are different, therefore, it is necessary to adjust the topological relationships of the dynamic model in real time, when solving the catapult dynamics of a carrier-based aircraft. In this paper, a dynamic model of the multibody system of the catapult take-off is established, and a variable topology solution is carried out for the dynamic model by adjusting dynamic augmentation equations; in addition, a dynamic analysis of a carrier-based aircraft catapult and take-off process is carried out. A catapult dynamics model and variable topology solution method were established, which solved the changes at the moment of the restraining rod separation, catapult rod separation, and catapult tackle during the aircraft catapult take-off. After the restraining rod was separated from the front landing gear, the catapult force was transmitted to the rear strut, which instantly increased the load of the rear strut by 238.5 kN. In addition, after the carrier-based aircraft reached the end of the catapult’s stroke, the catapult rod was separated from the catapult tow shuttle then unloaded, and the load of the rear strut was reduced from 486.2 kN to −20.3 kN. Under the protruding effect of the nose gear, the pitch angle of the carrier-based aircraft increased rapidly from −0.93° and reached 0.54° when the carrier-based aircraft rushed out of the deck.

Research on Recognition of Topological Relations Between Residential Lines in Low Voltage Station Area Based on Correlation Analysis Algorithm and Probabilistic Decision Method

In order to solve the problems of low accuracy and incomprehensive recognition of the topological relationship between households in the station area and the incomplete recognition results in traditional methods, a method for identifying topological relationships between household changes in low-voltage stations based on correlation analysis algorithm and probabilistic decision method is proposed. The BIRCH method is used to cluster the topological relationship characteristics of the household line changes in the low-voltage station area, and the topological relationship characteristics are obtained through clustering parameter initialization, clustering implementation and clustering evaluation, and the user phases in the topological relationship are identified according to the feature clustering results. The correlation analysis method is used to analyze the similarity of the voltage sequence of the points to be identified and the comprehensive similarity of all the faults of the target distribution transformer and the auxiliary distribution transformer, and set a similarity threshold to determine whether the points to be identified belong to the same station area. Finally, based on the probabilistic decision-making method, the identification of the topological relationship of the low-voltage station area household line change is completed. The experimental results show that this method can not only identify the topological relationship of single distribution transformer outage, but also identify the topological relationship of multiple distribution transformer outage. The accuracy of the identification result is high, and the identification loss function is low, which indicates that the identification result of this method is reliable and comprehensive.

Assessment and Visualization of OSM Consistency for European Cities

Volunteered Geographic Information (VGI) is a widely used data source in various fields and services, such as environmental monitoring, disaster and crisis management, SDI, and mapping. Quality is a critical factor for the usability of VGI. This study focuses on evaluating logical consistency based on the topological relationships between geographic features while considering semantics. It addresses internal (i.e., between thematic layers) and external (i.e., between specific features from different thematic layers) logical consistency. Attribute completeness is computed to support the use of semantics. A tool for assessing the consistency and attribute completeness is designed and implemented in the ArcGIS environment. An open-source web mapping application informs users about VGI consistency with multiscale visualization and indices. Data from OpenStreetMap (OSM), one of the most popular collaborative projects, are evaluated for six European cities: Athens, Berlin, Paris, Utrecht, Vienna, and Zurich. The case study uses OSM-derived data, downloaded from Geofabrik and organized into thematic layers. OSM’s consistency is evaluated and visualized at the regional, city, and feature levels. The results are discussed and conclusions on attribute completeness and consistency are derived.

Animal, Body, Data: Starling Murmurations and the Dynamic of Becoming In-formation

The aim of this article is to demonstrate that data modelling is becoming a crucial, if not dominant, vector for our understanding of animal populations and is consequential for how we study the affective relations between individual bodies and the communities to which they belong. It takes up the relationship between animal, body and data, following the datafication of starling murmurations, to explore the topological relationships between nature, culture and science. The case study thus embodies a data journey, invoking the tactics claimed by social or natural scientists, who generated recent discoveries in starling murmurations, including their topological expansions and contractions. The article concludes with thoughts and suggestions for further research on animal/data entanglement, and threads the concept of databodiment throughout, as a necessary dynamic for the formation and maintenance of communities.

Refining Convergent Rate Analysis with Topology in Mammalian Longevity and Marine Transitions

The quest to map the genetic foundations of phenotypes has been empowered by the modern diversity, quality, and availability of genomic resources. Despite these expanding resources, the abundance of variation within lineages makes the association of genetic change to specific phenotypes improbable. Drawing such connections requires an a priori means of isolating the associated changes from background genomic variation. Evolution may provide these means via convergence; i.e., the shared variation that may result from replicate evolutionary experiments across independent trait occurrences. To leverage these opportunities, we developed TRACCER: Topologically Ranked Analysis of Convergence via Comparative Evolutionary Rates. As compared to current methods, this software empowers rate convergence analysis by factoring in topological relationships, because variation between phylogenetically proximate trait changes is more likely to be facilitating the trait. Pairwise comparisons are performed not with singular branches, but in reference to their most recent common ancestors. This ensures that comparisons represent identical genetic contexts and timeframes while obviating the problematic requirement of assigning ancestral states. We applied TRACCER to two case studies: marine mammal transitions, an unambiguous trait which has independently evolved three times, as well as the evolution of mammalian longevity, a less delineated trait but with more instances to compare. TRACCER, by factoring in topology, identifies highly significant, convergent genetic signals in these test cases, with important incongruities and statistical resolution when compared to existing convergence approaches. These improvements in sensitivity and specificity generate refined targets for downstream analysis of convergent evolution and identification of genotype-phenotype relationships.

Distortion Effects in Equal Area Unit Maps

Maps that correctly represent the geographic size and shape of regions, taking into account scaling and generalization, have the disadvantage that small regions can easily be overlooked or not seen at all. Hence, for some map use tasks where small regions are of importance, alternative map types are needed. One option is the so-called equal area unit maps (EAUMs), where every enumeration unit has the same area size, possibly also the same basic shape such as squares or hexagons. The geometrical distortion of EAUMs, however, leads to a more difficult search for regions as well as a falsification of topological relationships and spatial patterns. To describe these distortions, a set of analytical measures is proposed. But it turns out that the expressiveness of these measures is rather limited. To better understand and to model the influence of distortions, two user studies were conducted. The study on the search in EAUMs (also with the aim of reconstruct the search strategies of the users) revealed how important it is to consider the local topology (e.g. corner or border positions of regions) during the generation process. With regard to pattern identification, it could be shown that EAUMs significantly increase the detection rate of local extreme values. On the other hand, global lateral gradients or geostatistical hot spots often get blurred or even lost. As a consequence, a task-oriented selection of map types and further developments are recommended.