اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدThree-Dimensional Spectral Classification of Low-Metallicity Stars Using Artificial Neural Networks
119
0
0.0
(
0
)
تأليف
Shawn Snider
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
We explore the application of artificial neural networks (ANNs) for the estimation of atmospheric parameters (Teff, logg, and [Fe/H]) for Galactic F- and G-type stars. The ANNs are fed with medium-resolution (~ 1-2 A) non flux-calibrated spectroscopic observations. From a sample of 279 stars with previous high-resolution determinations of metallicity, and a set of (external) estimates of temperature and surface gravity, our ANNs are able to predict Teff with an accuracy of ~ 135-150 K over the range 4250 <= Teff <= 6500 K, logg with an accuracy of ~ 0.25-0.30 dex over the range 1.0 <= logg <= 5.0 dex, and [Fe/H] with an accuracy ~ 0.15-0.20 dex over the range -4.0 <= [Fe/H] <= +0.3. Such accuracies are competitive with the results obtained by fine analysis of high-resolution spectra. It is noteworthy that the ANNs are able to obtain these results without consideration of photometric information for these stars. We have also explored the impact of the signal-to-noise ratio (S/N) on the behavior of ANNs, and conclude that, when analyzed with ANNs trained on spectra of commensurate S/N, it is possible to extract physical parameter estimates of similar accuracy with stellar spectra having S/N as low as 13. Taken together, these results indicate that the ANN approach should be of primary importance for use in present and future large-scale spectroscopic surveys.
قيم البحث
اقرأ أيضاً
We present a new methodology for the estimation of stellar atmospheric parameters from narrow- and intermediate-band photometry of the Javalambre Photometric Local Universe Survey (J-PLUS), and propose a method for target pre-selection of low-metalli
city stars for follow-up spectroscopic studies. Photometric metallicity estimates for stars in the globular cluster M15 are determined using this method. By development of a neural-network-based photometry pipeline, we aim to produce estimates of effective temperature, $T_{rm eff}$, and metallicity, [Fe/H], for a large subset of stars in the J-PLUS footprint. The Stellar Photometric Index Network Explorer, SPHINX, is developed to produce estimates of $T_{rm eff}$ and [Fe/H], after training on a combination of J-PLUS photometric inputs and synthetic magnitudes computed for medium-resolution (R ~ 2000) spectra of the Sloan Digital Sky Survey. This methodology is applied to J-PLUS photometry of the globular cluster M15. Effective temperature estimates made with J-PLUS Early Data Release photometry exhibit low scatter, sigma($T_{rm eff}$) = 91 K, over the temperature range 4500 < $T_{rm eff}$ (K) < 8500. For stars from the J-PLUS First Data Release with 4500 < $T_{rm eff}$ (K) < 6200, 85 $pm$ 3% of stars known to have [Fe/H] <-2.0 are recovered by SPHINX. A mean metallicity of [Fe/H]=-2.32 $pm$ 0.01, with a residual spread of 0.3 dex, is determined for M15 using J-PLUS photometry of 664 likely cluster members. We confirm the performance of SPHINX within the ranges specified, and verify its utility as a stand-alone tool for photometric estimation of effective temperature and metallicity, and for pre-selection of metal-poor spectroscopic targets.
New generation large-aperture telescopes, multi-object spectrographs, and large format detectors are making it possible to acquire very large samples of stellar spectra rapidly. In this context, traditional star-by-star spectroscopic analysis are no
longer practical. New tools are required that are capable of extracting quickly and with reasonable accuracy important basic stellar parameters coded in the spectra. Recent analyses of Artificial Neural Networks (ANNs) applied to the classification of astronomical spectra have demonstrated the ability of this concept to derive estimates of temperature and luminosity. We have adapted the back-propagation ANN technique developed by von Hippel et al. (1994) to predict effective temperatures, gravities and overall metallicities from spectra with resolving power ~ 2000 and low signal-to-noise ratio. We show that ANN techniques are very effective in executing a three-parameter (Teff,log g,[Fe/H]) stellar classification. The preliminary results show that the technique is even capable of identifying outliers from the training sample.
This paper investigates the problem of classification of unmanned aerial vehicles (UAVs) from radio frequency (RF) fingerprints at the low signal-to-noise ratio (SNR) regime. We use convolutional neural networks (CNNs) trained with both RF time-serie
s images and the spectrograms of 15 different off-the-shelf drone controller RF signals. When using time-series signal images, the CNN extracts features from the signal transient and envelope. As the SNR decreases, this approach fails dramatically because the information in the transient is lost in the noise, and the envelope is distorted heavily. In contrast to time-series representation of the RF signals, with spectrograms, it is possible to focus only on the desired frequency interval, i.e., 2.4 GHz ISM band, and filter out any other signal component outside of this band. These advantages provide a notable performance improvement over the time-series signals-based methods. To further increase the classification accuracy of the spectrogram-based CNN, we denoise the spectrogram images by truncating them to a limited spectral density interval. Creating a single model using spectrogram images of noisy signals and tuning the CNN model parameters, we achieve a classification accuracy varying from 92% to 100% for an SNR range from -10 dB to 30 dB, which significantly outperforms the existing approaches to our best knowledge.
The ability to accurately perceive whether a speaker is asking a question or is making a statement is crucial for any successful interaction. However, learning and classifying tonal patterns has been a challenging task for automatic speech recognitio
n and for models of tonal representation, as tonal contours are characterized by significant variation. This paper provides a classification model of Cypriot Greek questions and statements. We evaluate two state-of-the-art network architectures: a Long Short-Term Memory (LSTM) network and a convolutional network (ConvNet). The ConvNet outperforms the LSTM in the classification task and exhibited an excellent performance with 95% classification accuracy.
With the dual aims of enlarging the list of extremely metal-poor stars identified in the Galaxy, and boosting the numbers of moderately metal-deficient stars in directions that sample the rotational properties of the thick disk, we have used the 2.5m
Isaac Newton Telescope and the Intermediate Dispersion Spectrograph to carry out a survey of brighter (primarily northern hemisphere) metal-poor candidates selected from the HK objective-prism/interference-filter survey of Beers and collaborators. Over the course of only three observing runs (15 nights) we have obtained medium-resolution (resolving power ~ 2000) spectra for 1203 objects (V ~ 11-15). Spectral absorption-line indices and radial velocities have been measured for all of the candidates. Metallicities, quantified by [Fe/H], and intrinsic (B-V)o colors have been estimated for 731 stars with effective temperatures cooler than roughly 6500 K, making use of artificial neural networks (ANNs), trained with spectral indices. We show that this method performs as well as a previously explored Ca II K calibration technique, yet it presents some practical advantages. Among the candidates in our sample, we identify 195 stars with [Fe/H] <= -1.0, 67 stars with [Fe/H] <= -2.0, and 12 new stars with [Fe/H] <= -3.0. Although the EFECTIVE YIELD of metal-poor stars in our sample is not as large as previous HK survey follow-up programs, the rate of discovery per unit of telescope time is quite high.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
