Browsing by Author "Bauer, Franz Erik"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Alert Classification for the ALeRCE Broker System: The Anomaly Detector
    (IOP Publishing Ltd, 2023) Pérez-Carrasco, Manuel; Cabrera-Vives, Guillermo; Hernández-García, Lorena; Forster, F.; Sanchez-Saez, Paula; Muñoz Arancibia, Alejandra M.; Arredondo, Javier; Astorga, Nicolas; Bauer, Franz Erik; Bayo, Amelia; Catelan, Marcio; Dastidar, Raya; Estevez, P. A.; Lira, Paulina; Pignata, Giuliano
    Astronomical broker systems, such as Automatic Learning for the Rapid Classification of Events (ALeRCE), are currently analyzing hundreds of thousands of alerts per night, opening up an opportunity to automatically detect anomalous unknown sources. In this work, we present the ALeRCE anomaly detector, composed of three outlier detection algorithms that aim to find transient, periodic, and stochastic anomalous sources within the Zwicky Transient Facility data stream. Our experimental framework consists of cross-validating six anomaly detection algorithms for each of these three classes using the ALeRCE light-curve features. Following the ALeRCE taxonomy, we consider four transient subclasses, five stochastic subclasses, and six periodic subclasses. We evaluate each algorithm by considering each subclass as the anomaly class. For transient and periodic sources the best performance is obtained by a modified version of the deep support vector data description neural network, while for stochastic sources the best results are obtained by calculating the reconstruction error of an autoencoder neural network. Including a visual inspection step for the 10 most promising candidates for each of the 15 ALeRCE subclasses, we detect 31 bogus candidates (i.e., those with photometry or processing issues) and seven potential astrophysical outliers that require follow-up observations for further analysis.
  • Thumbnail Image
    Item
    BASS. XXXIV. A Catalog of the Nuclear Millimeter-wave Continuum Emission Properties of AGNs Constrained on Scales ≤ 100-200 pc
    (IOP Publishing Ltd, 2023) Kawamuro, Taiki; Ricci, Claudio; Mushotzky, Richard F.; Imanishi, Masatoshi; Bauer, Franz Erik; Ricci, Federica; Koss, Michael J.; Privon, George C.; Trakhtenbrot, Benny; Izumi, Takuma; Ichikawa, Kohei; Rojas, Alejandra F.; Smith, Krista Lynne; Shimizu, Taro; Oh, Kyuseok; den Brok, Jakob S.; Baba, Shunsuke; Balokovic, Mislav; Chang, Chin-Shin; Kakkad, Darshan; Pfeifle, Ryan W.; Temple, Matthew J.; Ueda, Yoshihiro; Harrison, Fiona; Powell, Meredith C.; Stern, Daniel; Urry, Meg; Sanders, David B.
    We present a catalog of the millimeter-wave (mm-wave) continuum properties of 98 nearby (z < 0.05) active galactic nuclei (AGNs) selected from the 70 month Swift/BAT hard-X-ray catalog that have precisely determined X-ray spectral properties and subarcsecond-resolution Atacama Large Millimeter/submillimeter Array Band 6 (211-275 GHz) observations as of 2021 April. Due to the hard-X-ray (>10 keV) selection, the sample is nearly unbiased for obscured systems at least up to Compton-thick-level obscuration, and provides the largest number of AGNs with high-physical-resolution mm-wave data (less than or similar to 100-200 pc). Our catalog reports emission peak coordinates, spectral indices, and peak fluxes and luminosities at 1.3 mm (230 GHz). Additionally, high-resolution mm-wave images are provided. Using the images and creating radial surface brightness profiles of mm-wave emission, we identify emission extending from the central sources and isolated blob-like emission. Flags indicating the presence of these emission features are tabulated. Among 90 AGNs with significant detections of nuclear emission, 37 AGNs (approximate to 41%) appear to have both or one of extended or blob-like components. We, in particular, investigate AGNs that show well-resolved mm-wave components and find that these seem to have a variety of origins (i.e., a jet, radio lobes, a secondary AGN, stellar clusters, a narrow-line region, galaxy disk, active star formation regions, or AGN-driven outflows), and some components have currently unclear origins.
  • Thumbnail Image
    Item
    Multi-Class Deep SVDD: Anomaly Detection Approach in Astronomy with Distinct Inlier Categories
    (Cornell Univ., 2023) Perez-Carrasco, Manuel; Bauer, Franz Erik; Hernandez-Garcia, Lorena; Forster, Francisco; Sanchez-Saez, Paula; Arancibia, Alejandra Munoz; Astorga, Nicolas; Bayo, Amelia; Cadiz-Leyton, Martina; Catelan, Marcio; Estevez, P. A.
    With the increasing volume of astronomical data generated by modern survey telescopes, automated pipelines and machine learning techniques have become crucial for analyzing and extracting knowledge from these datasets. Anomaly detection, i.e. the task of identifying irregular or unexpected patterns in the data, is a complex challenge in astronomy. In this paper, we propose Multi-Class Deep Support Vector Data Description (MCDSVDD), an extension of the state-of-the-art anomaly detection algorithm One-Class Deep SVDD, specifically designed to handle different inlier categories with distinct data distributions. MCDSVDD uses a neural network to map the data into hyperspheres, where each hypersphere represents a specific inlier category. The distance of each sample from the centers of these hyperspheres determines the anomaly score. We evaluate the effectiveness of MCDSVDD by comparing its performance with several anomaly detection algorithms on a large dataset of astronomical light-curves obtained from the Zwicky Transient Facility. Our results demonstrate the efficacy of MCDSVDD in detecting anomalous sources while leveraging the presence of different inlier categories. The code and the data needed to reproduce our results are publicly available at