A Survey of Few-Shot Learning: An Effective Method for Intrusion Detection

Few-shot learning (FSL) is a core topic in the domain of machine learning (ML), in which the focus is on the use of small datasets to train the model. In recent years, there have been many important data-driven ML applications for intrusion detection. Despite these great achievements, however, gathering a large amount of reliable data remains expensive and time-consuming, or even impossible. In this regard, FSL has been shown to have advantages in terms of processing small, abnormal data samples in the huge application space of intrusion detection. FSL can improve ML for scarce data at three levels: the data, the model, and the algorithm levels. Previous knowledge plays an important role in all three approaches. Many promising methods such as data enrichment, the graph neural network model, and multitask learning have also been developed. In this paper, we present a comprehensive review of the latest research progress in the area of FSL. We first introduce the theoretical background to ML and FSL and then describe the general features, advantages, and main methods of FSL. FSL methods such as embedded learning, multitask learning, and generative models are applied to intrusion detection to improve the detection accuracy effectively. Then, the application of FSL to intrusion detection is reviewed in detail, including enriching the dataset by extracting intermediate features, using graph embedding and meta-learning methods to improve the model. Finally, the difficulties of this approach and its prospects for development in the field of intrusion detection are identified based on the previous discussion.

LiDAR Data Enrichment by Fusing Spatial and Temporal Adjacent Frames

In autonomous driving scenarios, the point cloud generated by LiDAR is usually considered as an accurate but sparse representation. In order to enrich the LiDAR point cloud, this paper proposes a new technique that combines spatial adjacent frames and temporal adjacent frames. To eliminate the “ghost” artifacts caused by moving objects, a moving point identification algorithm is introduced that employs the comparison between range images. Experiments are performed on the publicly available Semantic KITTI dataset. Experimental results show that the proposed method outperforms most of the previous approaches. Compared with these previous works, the proposed method is the only method that can run in real-time for online usage.

An Extensible Data Enrichment Scheme for Providing Intelligent Services in Internet of Things Environments

Recent technologies in the Internet of Things (IoT) environment aim to provide intelligent services to users. Intelligent services can be managed and executed by systems that handle context information sets. Handling intelligent services leads to three major considerations: objects in the real world that should be described as metadata, a data enrichment procedure from sensing values for representing states, and controlling functionalities to manage services. In this study, an extensible data-enrichment scheme is proposed. The proposed scheme provides a way to describe profiles, data abstraction procedures, and functionalities that support the building of context information sets derived from raw datasets in the manner of a semantic web stack. Finally, data enrichment will help any system that uses context information by providing improved, understandable, and readable datasets to the service developers or the systems themselves.

Unexpected role of SIX1 variants in craniosynostosis: expanding the phenotype of SIX1-related disorders

BackgroundPathogenic heterozygous SIX1 variants (predominantly missense) occur in branchio-otic syndrome (BOS), but an association with craniosynostosis has not been reported.MethodsWe investigated probands with craniosynostosis of unknown cause using whole exome/genome (n=628) or RNA (n=386) sequencing, and performed targeted resequencing of SIX1 in 615 additional patients. Expression of SIX1 protein in embryonic cranial sutures was examined in the Six1nLacZ/+ reporter mouse.ResultsFrom 1629 unrelated cases with craniosynostosis we identified seven different SIX1 variants (three missense, including two de novo mutations, and four nonsense, one of which was also present in an affected twin). Compared with population data, enrichment of SIX1 loss-of-function variants was highly significant (p=0.00003). All individuals with craniosynostosis had sagittal suture fusion; additionally four had bilambdoid synostosis. Associated BOS features were often attenuated; some carrier relatives appeared non-penetrant. SIX1 is expressed in a layer basal to the calvaria, likely corresponding to the dura mater, and in the mid-sagittal mesenchyme.ConclusionCraniosynostosis is associated with heterozygous SIX1 variants, with possible enrichment of loss-of-function variants compared with classical BOS. We recommend screening of SIX1 in craniosynostosis, particularly when sagittal±lambdoid synostosis and/or any BOS phenotypes are present. These findings highlight the role of SIX1 in cranial suture homeostasis.

Speed Limit Induced CO2 Reduction on Motorways: Enhancing Discussion Transparency through Data Enrichment of Road Networks

Considering climate change, recent political debates often focus on measures to reduce CO2 emissions. One key component is the reduction of emissions produced by motorized vehicles. Since the amount of emission directly correlates to the velocity of a vehicle via energy consumption factors, a general speed limit is often proposed. This article presents a methodology to combine openly available topology data of road networks from OpenStreetMap (OSM) with pay-per-use API traffic data from TomTom to evaluate such measures transparently by analyzing historical real-world circumstances. From our exemplary case study of the German motorway network, we derive that most parts of the motorway network on average do not reach their maximum allowed speed throughout the day due to traffic, construction sites and general road utilization by network participants. Nonetheless our findings prove that the introduction of a speed limit of 120 km per hour on the German autobahn would restrict 50.74% of network flow kilometers for a CO2 reduction of 7.43% compared to the unrestricted state.