Search

In line with the tradition of employing nuclear reactors for neutrino physics research, we present potential experiments that could be undertaken using the new Brazilian neutron source, the Brazilian Multipurpose Reactor (RMB). This upcoming facility has clearly defined objectives, including the production of radioisotopes for the healthcare system, material irradiation testing, and research using neutron beams to investigate the micro and mesostructures of various materials. However, we anticipate that the RMB will generate up to $3,500$ antineutrinos daily, potentially expanding the facility's scope to include neutrino physics experiments involving beta and inverse beta reactions. The modern operational conditions of the RMB and its location in the Iper\'o-Brazil region may also enable neutrino oscillation experiments and the determination of the $\theta_{13}$ parameter, as well as support safeguards activities by the International Atomic Energy Agency (IAEA), allowing for the monitoring of nuclear fuel burning cycles and isotope fractions. We discuss the potential utilization of the RMB as an antineutrino source, proposing experiments and addressing associated technical limitations., Comment: 13 pages, 5 figures, some corrections, and references added

Smart-parking solutions use sensors, cameras, and data analysis to improve parking efficiency and reduce traffic congestion. Computer vision-based methods have been used extensively in recent years to tackle the problem of parking lot management, but most of the works assume that the parking spots are manually labeled, impacting the cost and feasibility of deployment. To fill this gap, this work presents an automatic parking space detection method, which receives a sequence of images of a parking lot and returns a list of coordinates identifying the detected parking spaces. The proposed method employs instance segmentation to identify cars and, using vehicle occurrence, generate a heat map of parking spaces. The results using twelve different subsets from the PKLot and CNRPark-EXT parking lot datasets show that the method achieved an AP25 score up to 95.60\% and AP50 score up to 79.90\%., Comment: Accepted for presentation at the 2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC 2023)

Neutron imaging is a non-destructive technique for analyzing a wide class of samples, such as archaeological or industrial material structures. In recent decades, technological advances have had a great impact on the neutron imaging technique, which has meant an evolution from simple radiographs using films (2D) to modern tomography systems with digital processing (3D). The 5 MW research nuclear reactor IEA-R1, which is located at the Instituto de Pesquisas Energ\'eticas e Nucleares (IPEN) in Brazil, possesses a neutron imaging instrument with $1.0 \times 10^{6}$ $n/cm^{2}s$ in the sample position. IEA-R1 is over 60 years old and the future of neutron science in Brazil, including imaging, will be expanded to a new facility called the Brazilian Multipurpose Reactor (RMB, Portuguese acronym), which will be built soon. The new reactor will house a suite of instruments at the Neutron National Laboratory, including the neutron imaging facility, viz., Neinei. Inspired by recent author's works, we model the Neinei instrument through stochastic Monte Carlo simulations. We investigate the sensitivity of the neutron imaging technique parameter ($L/D$ ratio) with the neutron flux, and the results are compared to data from the Neutra (PSI), Antares (FRM II), BT2 (NIST) and DINGO (OPAL) instruments. The results are promising and provide avenues for future improvements., Comment: 12 pages, 11 figures. Version accepted in the Nuclear Engineering and Design (2024)

This paper presents a novel approach combining convolutional layers (CLs) and large-margin metric learning for training supervised models on small datasets for texture classification. The core of such an approach is a loss function that computes the distances between instances of interest and support vectors. The objective is to update the weights of CLs iteratively to learn a representation with a large margin between classes. Each iteration results in a large-margin discriminant model represented by support vectors based on such a representation. The advantage of the proposed approach w.r.t. convolutional neural networks (CNNs) is two-fold. First, it allows representation learning with a small amount of data due to the reduced number of parameters compared to an equivalent CNN. Second, it has a low training cost since the backpropagation considers only support vectors. The experimental results on texture and histopathologic image datasets have shown that the proposed approach achieves competitive accuracy with lower computational cost and faster convergence when compared to equivalent CNNs., Comment: 7 pages

Malware is a major threat to computer systems and imposes many challenges to cyber security. Targeted threats, such as ransomware, cause millions of dollars in losses every year. The constant increase of malware infections has been motivating popular antiviruses (AVs) to develop dedicated detection strategies, which include meticulously crafted machine learning (ML) pipelines. However, malware developers unceasingly change their samples' features to bypass detection. This constant evolution of malware samples causes changes to the data distribution (i.e., concept drifts) that directly affect ML model detection rates, something not considered in the majority of the literature work. In this work, we evaluate the impact of concept drift on malware classifiers for two Android datasets: DREBIN (about 130K apps) and a subset of AndroZoo (about 285K apps). We used these datasets to train an Adaptive Random Forest (ARF) classifier, as well as a Stochastic Gradient Descent (SGD) classifier. We also ordered all datasets samples using their VirusTotal submission timestamp and then extracted features from their textual attributes using two algorithms (Word2Vec and TF-IDF). Then, we conducted experiments comparing both feature extractors, classifiers, as well as four drift detectors (DDM, EDDM, ADWIN, and KSWIN) to determine the best approach for real environments. Finally, we compare some possible approaches to mitigate concept drift and propose a novel data stream pipeline that updates both the classifier and the feature extractor. To do so, we conducted a longitudinal evaluation by (i) classifying malware samples collected over nine years (2009-2018), (ii) reviewing concept drift detection algorithms to attest its pervasiveness, (iii) comparing distinct ML approaches to mitigate the issue, and (iv) proposing an ML data stream pipeline that outperformed literature approaches.

The Brazilian Multipurpose Reactor (RMB) is a fundamental project that aims to turn Brazil into a self-sufficient country in the production of radioisotopes and radiopharmaceuticals to supply the Unified Health System (SUS) as much as the private institutions. In addition, the RMB project describes other applications as irradiation and testing of nuclear fuels and structural material analysis, for instance. There are many techniques in the project to study structural aspects of materials, where neutron diffraction represents one of the priorities for implementation. This technique will take place mainly on two diffractometers on Thermal Neutron Guide 1 (TG1), namely Araponga, a high-resolution diffractometer, and Flautim, a high-intensity diffractometer. In this work, we study the performance of the Araponga diffractometer through McStas simulations with input produced by the MCNP code of the RMB core. We investigate the neutron flux values considering a state-of-art high-resolution diffractometer, and the results are promising since some simulated scenarios present values compatible with high-intensity devices., Comment: 10 pages, 2 figures

The equilibrium transport properties of an elementary nanostructured device with side-coupled geometry are computed and related to universal functions. The computation relies on a real-space formulation of the numerical renormalization-group (NRG) procedure. The real-space construction, dubbed eNRG, is more straightforward than the NRG discretization and allows more faithful description of the coupling between quantum dots and conduction states. The procedure is applied to an Anderson-model description of a quantum wire side-coupled to a single quantum dot. A gate potential controls the dot occupation. In the Kondo regime, the electrical conductance through this device is known to map linearly onto a universal function of the temperature scaled by the Kondo temperature. Here, the energy moments from which the Seebeck coefficient and the thermal conductance can be computed are shown to map linearly onto universal functions also. The moments and transport properties computed by the eNRG procedure are shown to agree very well with these analytical developments. Algorithms facilitating comparison with experimental results are discussed. As an illustration, one of the algorithms is applied to thermal dependence of the thermopower measured by K\"{o}hler [PhD Thesis, TUD, Dresden, 2007] in Lu$_{0.9}$Yb$_{0.1}$Rh$_{2}$Si$_{2}$., Comment: 21 pages, 11 figures; submitted to Phys. Rev. B

During the last decades, neutron beam transportation has been a well-known and established subject for designing proper neutron guides. However, sometimes unusual adaptation or adjustments are required out of original projects and after operation beginning of facilities. Inter-center transferring of instrument locations also requires a new approach that is not necessarily described in the literature. Inside these situations, the use of S-shaped guides has not been fully discussed in the literature. From a geometrical analysis, we develop a formalism of construction of a minimal S-shaped guide by only considering the exclusion of the Line-of-Sight. We study this guide model through the wavelength cutoff and the neutron transport efficiency analysis. Here, Monte Carlo simulations using MCSTAS software are applied. By intending to optimize these guide systems, simulations of this study also consider scenarios that have different supermirrors. A formalism to determine wavelength cutoff for unique and variable index guide systems is also developed. Simulation results show a good agreement between theoretical and simulated wavelength cutoff values. In addition, we have found specific configurations that combine efficient neutron transport and lower index values on the convex surfaces of curved guides that form the S-shaped guide., Comment: 25 pages, Preprint submitted to Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment

The availability of labelled data is one of the main limitations in machine learning. We can alleviate this using weak supervision: a framework that uses expert-defined rules $\boldsymbol{\lambda}$ to estimate probabilistic labels $p(y|\boldsymbol{\lambda})$ for the entire data set. These rules, however, are dependent on what experts know about the problem, and hence may be inaccurate or may fail to capture important parts of the problem-space. To mitigate this, we propose Active WeaSuL: an approach that incorporates active learning into weak supervision. In Active WeaSuL, experts do not only define rules, but they also iteratively provide the true label for a small set of points where the weak supervision model is most likely to be mistaken, which are then used to better estimate the probabilistic labels. In this way, the weak labels provide a warm start, which active learning then improves upon. We make two contributions: 1) a modification of the weak supervision loss function, such that the expert-labelled data inform and improve the combination of weak labels; and 2) the maxKL divergence sampling strategy, which determines for which data points expert labelling is most beneficial. Our experiments show that when the budget for labelling data is limited (e.g. $\leq 60$ data points), Active WeaSuL outperforms weak supervision, active learning, and competing strategies, with only a handful of labelled data points. This makes Active WeaSuL ideal for situations where obtaining labelled data is difficult., Comment: Accepted to the ICLR 2021 Workshop on Weakly Supervised Learning