اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدHow to Find More Supernovae with Less Work: Object Classification Techniques for Difference Imaging
365
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
We present the results of applying new object classification techniques to difference images in the context of the Nearby Supernova Factory supernova search. Most current supernova searches subtract reference images from new images, identify objects in these difference images, and apply simple threshold cuts on parameters such as statistical significance, shape, and motion to reject objects such as cosmic rays, asteroids, and subtraction artifacts. Although most static objects subtract cleanly, even a very low false positive detection rate can lead to hundreds of non-supernova candidates which must be vetted by human inspection before triggering additional followup. In comparison to simple threshold cuts, more sophisticated methods such as Boosted Decision Trees, Random Forests, and Support Vector Machines provide dramatically better object discrimination. At the Nearby Supernova Factory, we reduced the number of non-supernova candidates by a factor of 10 while increasing our supernova identification efficiency. Methods such as these will be crucial for maintaining a reasonable false positive rate in the automated transient alert pipelines of upcoming projects such as PanSTARRS and LSST.
قيم البحث
اقرأ أيضاً
Type Ia supernovae (SNe Ia) that are multiply imaged by gravitational lensing can extend the SN Ia Hubble diagram to very high redshifts $(zgtrsim 2)$, probe potential SN Ia evolution, and deliver high-precision constraints on $H_0$, $w$, and $Omega_
m$ via time delays. However, only one, iPTF16geu, has been found to date, and many more are needed to achieve these goals. To increase the multiply imaged SN Ia discovery rate, we present a simple algorithm for identifying gravitationally lensed SN Ia candidates in cadenced, wide-field optical imaging surveys. The technique is to look for supernovae that appear to be hosted by elliptical galaxies, but that have absolute magnitudes implied by the apparent hosts photometric redshifts that are far brighter than the absolute magnitudes of normal SNe Ia (the brightest type of supernovae found in elliptical galaxies). Importantly, this purely photometric method does not require the ability to resolve the lensed images for discovery. AGN, the primary sources of contamination that affect the method, can be controlled using catalog cross-matches and color cuts. Highly magnified core-collapse supernovae will also be discovered as a byproduct of the method. Using a Monte Carlo simulation, we forecast that LSST can discover up to 500 multiply imaged SNe Ia using this technique in a 10-year $z$-band search, more than an order of magnitude improvement over previous estimates (Oguri & Marshall 2010). We also predict that ZTF should find up to 10 multiply imaged SNe Ia using this technique in a 3-year $R$-band search---despite the fact that this survey will not resolve a single system.
Text classification is a critical research topic with broad applications in natural language processing. Recently, graph neural networks (GNNs) have received increasing attention in the research community and demonstrated their promising results on t
his canonical task. Despite the success, their performance could be largely jeopardized in practice since they are: (1) unable to capture high-order interaction between words; (2) inefficient to handle large datasets and new documents. To address those issues, in this paper, we propose a principled model -- hypergraph attention networks (HyperGAT), which can obtain more expressive power with less computational consumption for text representation learning. Extensive experiments on various benchmark datasets demonstrate the efficacy of the proposed approach on the text classification task.
Our main problem is to find finite topological spaces to within homeomorphism, given (also to within homeomorphism) the quotient-spaces obtained by identifying one point of the space with each one of the other points. In a previous version of this pa
per, our aim was to reconstruct a topological space from its quotient-spaces; but a reconstruction is not always possible either in the sense that several non-homeomorphic topological spaces yield the same quotient-spaces, or in the sense that no topological space yields an arbitrarily given family of quotient-spaces. In this version of the paper we present an algorithm that detects, and deals with, all these situations.
In recent years there has been an increasing trend in which data scientists and domain experts work together to tackle complex scientific questions. However, such collaborations often face challenges. In this paper, we aim to decipher this collaborat
ion complexity through a semi-structured interview study with 22 interviewees from teams of bio-medical scientists collaborating with data scientists. In the analysis, we adopt the Olsons four-dimensions framework proposed in Distance Matters to code interview transcripts. Our findings suggest that besides the glitches in the collaboration readiness, technology readiness, and coupling of work dimensions, the tensions that exist in the common ground building process influence the collaboration outcomes, and then persist in the actual collaboration process. In contrast to prior works general account of building a high level of common ground, the breakdowns of content common ground together with the strengthen of process common ground in this process is more beneficial for scientific discovery. We discuss why that is and what the design suggestions are, and conclude the paper with future directions and limitations.
An abdominal ultrasound examination, which is the most common ultrasound examination, requires substantial manual efforts to acquire standard abdominal organ views, annotate the views in texts, and record clinically relevant organ measurements. Hence
, automatic view classification and landmark detection of the organs can be instrumental to streamline the examination workflow. However, this is a challenging problem given not only the inherent difficulties from the ultrasound modality, e.g., low contrast and large variations, but also the heterogeneity across tasks, i.e., one classification task for all views, and then one landmark detection task for each relevant view. While convolutional neural networks (CNN) have demonstrated more promising outcomes on ultrasound image analytics than traditional machine learning approaches, it becomes impractical to deploy multiple networks (one for each task) due to the limited computational and memory resources on most existing ultrasound scanners. To overcome such limits, we propose a multi-task learning framework to handle all the tasks by a single network. This network is integrated to perform view classification and landmark detection simultaneously; it is also equipped with global convolutional kernels, coordinate constraints, and a conditional adversarial module to leverage the performances. In an experimental study based on 187,219 ultrasound images, with the proposed simplified approach we achieve (1) view classification accuracy better than the agreement between two clinical experts and (2) landmark-based measurement errors on par with inter-user variability. The multi-task approach also benefits from sharing the feature extraction during the training process across all tasks and, as a result, outperforms the approaches that address each task individually.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
