Fakultet for naturvitenskap og teknologi: Nye registreringer
Viser treff 161-180 av 4870
-
Mabnet: Master Assistant Buddy Network With Hybrid Learning for Image Retrieval
(Journal article; Tidsskriftartikkel; Peer reviewed, 2023-05-05)Image retrieval has garnered a growing interest in recent times. The current approaches are either supervised or self-supervised. These methods do not exploit the benefits of hybrid learning using both supervision and self-supervision. We present a novel Master Assistant Buddy Network (MAB-Net) for image retrieval which incorporates both the learning mechanisms. MABNet consists of master and assistant ... -
Aux-Drop: Handling Haphazard Inputs in Online Learning Using Auxiliary Dropouts
(Journal article; Tidsskriftartikkel; Peer reviewed, 2023)Many real-world applications based on online learning produce streaming data that is haphazard in nature, i.e., contains missing features, features becoming obsolete in time, the appearance of new features at later points in time and a lack of clarity on the total number of input features. These challenges make it hard to build a learnable system for such applications, and almost no work exists in ... -
Estimating the Space Debris Density Function using Radar Beam Park Measurements
(Journal article; Tidsskriftartikkel; Peer reviewed, 2023)Estimating the density of space debris as a function of orbital elements and size is crucial for determining the risk of collision of spacecraft. For objects 1 cm or larger in diameter, this information can be obtained using beam park observations made with powerful ground based radars. This study presents a novel technique for estimating the density of space debris as a function of orbital elements ... -
Radar size inference from statistics of RCS samples
(Journal article; Tidsskriftartikkel; Peer reviewed, 2023)Radar Beam-park experiments have been very successful in characterizing the distribution of space debris objects, both in terms of orbital parameters but also in terms of limiting the estimates of their radar cross section, or RCS. A recent paper \[0\] used observed range and range rates to refine orbit estimates by matching up the observed SNR curve to that predicted by simulations. This gives good ... -
Linear codes associated to symmetric determinantal varieties: Even rank case
(Journal article; Tidsskriftartikkel; Peer reviewed, 2023-06-20)We consider linear codes over a finite field Fq, for odd q, derived from determinantal varieties, obtained from symmetric matrices of bounded ranks. A formula for the weight of a codeword is derived. Using this formula, we have computed the minimum distance for the codes corresponding to matrices upperbounded by any fixed, even rank. A conjecture is proposed for the cases where the upper bound is ... -
Exceptionally simple super-PDE for F (4)
(Journal article; Tidsskriftartikkel; Peer reviewed, 2024-02-01)For the largest exceptional simple Lie superalgebra F(4), having dimension (24|16), we provide two explicit geometric realizations as supersymmetries, namely as the symmetry superalgebra of super-PDE systems of second and third order respectively. -
ODEs whose Symmetry Groups are not Fiber-Preserving
(Journal article; Tidsskriftartikkel; Peer reviewed, 2023)We observe that, up to conjugation, a majority of higher order ODEs (ordinary differential equations) and ODE systems have only fiber-preserving point symmetries. By exploiting Lie's classification of Lie algebras of vector fields, we describe all exceptions to this in the case of scalar ODEs and systems of ODEs on a pair of functions. The scalar ODEs whose symmetry algebras are not fiber preserving ... -
ProtoVAE: A Trustworthy Self-Explainable Prototypical Variational Model
(Journal article; Tidsskriftartikkel; Peer reviewed, 2022-10-15)The need for interpretable models has fostered the development of self-explainable classifiers. Prior approaches are either based on multi-stage optimization schemes, impacting the predictive performance of the model, or produce explanations that are not transparent, trustworthy or do not capture the diversity of the data. To address these shortcomings, we propose ProtoVAE, a variational autoencoder-based ... -
Assigning the unassigned: A signature-based classification of rDNA metabarcodes reveals new deep-sea diversity
(Journal article; Tidsskriftartikkel; Peer reviewed, 2024-02-29)Environmental DNA metabarcoding reveals a vast genetic diversity of marine eukaryotes. Yet, most of the metabarcoding data remain unassigned due to the paucity of reference databases. This is particularly true for the deep-sea meiofauna and eukaryotic microbiota, whose hidden diversity is largely unexplored. Here, we tackle this issue by using unique DNA signatures to classify unknown metabarcodes ... -
Towards Interpretable, Trustworthy and Reliable AI
(Doctoral thesis; Doktorgradsavhandling, 2024-03-15)<p>The field of artificial intelligence recently witnessed remarkable growth, leading to the development of complex deep learning models that perform exceptionally across various domains. However, these developments bring forth critical issues. Deep learning models are vulnerable to inheriting and potentially exacerbating biases present in their training data. Moreover, the complexity of these models ... -
Synthesis of thiol-appended gold and rhenium corroles as potential nanoconjugants for gold nanoparticles
(Master thesis; Mastergradsoppgave, 2023-06-10)Photodynamic therapy (PDT) today is an established treatment for a variety of cancers as well as various dermatological conditions, macular degeneration and bacterial and other infections. Treatment involves the administration of a photosensitizer, typically a porphyrin, followed by irradiation with light. The photoexcited sensitizer transfers its excess energy to oxygen in the affected tissue, ... -
Investigating the Effect of Pressure and UV Radiation on Antioxidant and UV Stabilizing Additives in Different Tire Particles
(Master thesis; Mastergradsoppgave, 2023-05-25)This thesis investigates the impact four different experimental marine conditions have on the extractability of tire additives meant to protect tires from UV radiation as well as some other common substances found in tires. The impact is evaluated between tire particle size and age to determine if the exposure affects them differently. This thesis finds that the effect of marine conditions on ... -
Global change research needs international collaboration
(Journal article; Tidsskriftartikkel; Peer reviewed, 2023-08-04)Tackling the grand challenges of global climate change for the sustainability of ecological and societal systems requires data and expertise from Russia, the world's largest country that has the longest Arctic shoreline and the largest forest biome, peatland and permafrost zones. Academic relations and scientific collaborations with Russian scholars and institutions must continue despite the ensuing ... -
Scenario Design, Data Measurement, and Analysis Approaches in Maritime Simulator Training: A Systematic Review
(Chapter; Bokkapittel, 2023-08-29)Developing objective assessment approach in maritime simulator training can be a highly challenging task due to the complexity of simulating realistic scenarios, capturing relevant performance indicators and establishing good assessment protocols. This study provides a synthesis of simulation scenario contexts, data collection tools, and data analysis approaches in published simulator training ... -
Betydningen av for- og etterarbeid for å fremme faglig forståelse under praktisk kjemiundervisning på ungdomstrinnet
(Master thesis; Mastergradsoppgave, 2023-05-30)Denne oppgaven er skrevet på bakgrunn av interesse og nysgjerrighet rundt praktiske undervisningsmåter. Oppgaven har som mål å undersøke hvordan ungdomsskolelærere benytter seg av praktiske aktiviteter og for- og etterarbeid for å fremme elevenes faglige forståelse. Studien er utført ved analyse av LISSI-studien sitt videomateriale av norsk naturfagundervisning på ungdomstrinnet. Analysen bestod ... -
Deep Learning Based Automatic Segmentation of Gas Flares in Single Beam Echo Sounder Data
(Mastergradsoppgave; Master thesis, 2024-01-18)This thesis introduces the first study of instance segmentation applied to gas flares in single beam echo sounder data. We develop a comprehensive dataset consisting of 1,414 images, featuring 5,142 segmented objects identified as gas flare. A key contribution is the adaptation of the Brier score specifically for instance segmentation. Further, we show how to adapt the Weighted Box Fusion (WBF) ... -
Depositional environments and source rock potential of some Upper Palaeozoic (Devonian) coals on Bjørnøya, Western Barents shelf
(Journal article; Tidsskriftartikkel; Peer reviewed, 2024-02-27)Upper Palaeozoic strata are potential sources for hydrocarbon plays on the Norwegian Barents Shelf. The island of Bjørnøya (Bear Island) is located on the western margin of the Barents Shelf and represents an exposed part of the Stappen High, which makes it an excellent location to investigate the source rock potential of this succession. Here, we investigate the organic geochemical composition and ... -
Protracted post-glacial hydrocarbon seepage in the Barents Sea revealed by U–Th dating of seep carbonates
(Journal article; Tidsskriftartikkel; Peer reviewed, 2024-03-01)The hydrocarbon seepage chronology during deglaciation across the formerly glaciated Barents Sea was established using uranium-thorium (U–Th) dating of seep carbonates. Seep carbonates were sampled with remotely operated vehicles (ROV) from the seafloor at three active hydrocarbon seeps (water depth 156–383 m), located in the north-west (Storfjordrenna), north-central (Storbanken High), and south-west ... -
Decomposable (5, 6)-solutions in eleven-dimensional supergravity
(Journal article; Tidsskriftartikkel; Peer reviewed, 2023-06-08)We present decomposable (5, 6)-solutions M<sup>1,4</sup>×M<sup>6</sup> in eleven-dimensional supergravity by solving the bosonic supergravity equations for a variety of non-trivial flux forms. Many of the bosonic backgrounds presented here are induced by various types of null flux forms on products of certain totally Ricci-isotropic Lorentzian Walker manifolds and Ricci-flat Riemannian manifolds. ... -
Weather-aware Wake-up of Sleeping Cyber-Physical IoT Nodes
(Chapter; Bokkapittel, 2023-09-27)Cyber-physical IoT nodes located in environments which are resource-constrained and physically hard to access, like the Arctic tundra, must achieve long operational lifetimes from a single battery and report data over data networks. The nodes sleep most of the time, and only wake up to perform mission tasks, including reporting data. However, networks can become unavailable, or have low bandwidth ...