Search

The Beijing-Arizona Sky Survey (BASS) Data Release 3 (DR3) catalogue was released in 2019, which contains the data from all BASS and the Mosaic z-band Legacy Survey (MzLS) observations during 2015 January and 2019 March, about 200 million sources. We cross-match BASS DR3 with spectral databases from the Sloan Digital Sky Survey (SDSS) and the Large Sky Area Multi-object Fiber Spectroscopic Telescope (LAMOST) to obtain the spectroscopic classes of known samples. Then, the samples are cross-matched with ALLWISE database. Based on optical and infrared information of the samples, we use the XGBoost algorithm to construct different classifiers, including binary classification and multiclass classification. The accuracy of these classifiers with the best input pattern is larger than 90.0 per cent. Finally, all selected sources in the BASS DR3 catalogue are classified by these classifiers. The classification label and probabilities for individual sources are assigned by different classifiers. When the predicted results by binary classification are the same as multiclass classification with optical and infrared information, the number of star, galaxy and quasar candidates is separately 12 375 838 (P_S>0.95), 18 606 073 (P_G>0.95) and 798 928 (P_Q>0.95). For these sources without infrared information, the predicted results can be as a reference. Those candidates may be taken as input catalogue of LAMOST, DESI or other projects for follow up observation. The classified result will be of great help and reference for future research of the BASS DR3 sources., Comment: 15 pages, 10 tables, 6 figures, accepted for publication in Monthly Notices of the Royal Astronomical Society Main Journal

The ESA's X-ray Multi-Mirror Mission (XMM-Newton) created a new, high quality version of the XMM-Newton serendipitous source catalogue, 4XMM-DR9, which provides a wealth of information for observed sources. The 4XMM-DR9 catalogue is correlated with the Sloan Digital Sky Survey (SDSS) DR12 photometric database and the ALLWISE database, then we get the X-ray sources with information from X-ray, optical and/or infrared bands, and obtain the XMM-WISE sample, the XMM-SDSS sample and the XMM-WISE-SDSS sample. Based on the large spectroscopic surveys of SDSS and the Large Sky Area Multi-object Fiber Spectroscopic Telescope (LAMOST), we cross-match the XMM-WISE-SDSS sample with those sources of known spectral classes, and obtain the known samples of stars, galaxies and quasars. The distribution of stars, galaxies and quasars as well as all spectral classes of stars in 2-d parameter spaces is presented. Various machine learning methods are applied on different samples from different bands. The better classified results are retained. For the sample from X-ray band, rotation forest classifier performs the best. For the sample from X-ray and infrared bands, a random forest algorithm outperforms all other methods. For the samples from X-ray, optical and/or infrared bands, LogitBoost classifier shows its superiority. Thus, all X-ray sources in the 4XMM-DR9 catalogue with different input patterns are classified by their respective models which are created by these best methods. Their membership and membership probabilities to individual X-ray sources are assigned. The classified result will be of great value for the further research of X-ray sources in greater detail., Comment: 11 pages, 4 figures, 10 tables, accepted by Monthly Notices of the Royal Astronomical Society

Spectrum denoising is an important procedure for large-scale spectroscopical surveys. This work proposes a novel stellar spectrum denoising method based on deep Bayesian modeling. The construction of our model includes a prior distribution for each stellar subclass, a spectrum generator and a flow-based noise model. Our method takes into account the noise correlation structure, and it is not susceptible to strong sky emission lines and cosmic rays. Moreover, it is able to naturally handle spectra with missing flux values without ad-hoc imputation. The proposed method is evaluated on real stellar spectra from the Sloan Digital Sky Survey (SDSS) with a comprehensive list of common stellar subclasses and compared to the standard denoising auto-encoder. Our denoising method demonstrates superior performance to the standard denoising auto-encoder, in respect of denoising quality and missing flux imputation. It may be potentially helpful in improving the accuracy of the classification and physical parameter measurement of stars when applying our method during data preprocessing., Comment: 20 pages, 15 Figures, accepted by Research in Astronomy and Astrophysics

We combine K-Nearest Neighbors (KNN) with genetic algorithm (GA) for photometric redshift estimation of quasars, short for GeneticKNN, which is a weighted KNN approach supported by GA. This approach has two improvements compared to KNN: one is the feature weighted by GA; another is that the predicted redshift is not the redshift average of K neighbors but the weighted average of median and mean of redshifts for K neighbors, i.e. $p\times z_{median} + (1-p)\times z_{mean}$. Based on the SDSS and SDSS-WISE quasar samples, we explore the performance of GeneticKNN for photometric redshift estimation, comparing with the other six traditional machine learning methods, i.e. Least absolute shrinkage and selection operator (LASSO), support vector regression (SVR), Multi Layer Perceptrons (MLP), XGBoost, KNN and random forest. KNN and random forest show their superiority. Considering the easy implementation of KNN, we make improvement on KNN as GeneticKNN and apply GeneticKNN on photometric redshift estimation of quasars. Finally the performance of GeneticKNN is better than that of LASSO, SVR, MLP, XGBoost, KNN and random forest for all cases. Moreover the accuracy is better with the additional WISE magnitudes for the same method., Comment: 17 pages, 4 tables, 6 figures

As data from IoT (Internet of Things) sensors become ubiquitous, state-of-the-art machine learning algorithms face many challenges on directly using sensor data. To overcome these challenges, methods must be designed to learn directly from sensors without manual annotations. This paper introduces Sensory Time-cue for Unsupervised Meta-learning (STUM). Different from traditional learning approaches that either heavily depend on labels or on time-independent feature extraction assumptions, such as Gaussian distribution features, the STUM system uses time relation of inputs to guide the feature space formation within and across modalities. The fact that STUM learns from a variety of small tasks may put this method in the camp of Meta-Learning. Different from existing Meta-Learning approaches, STUM learning tasks are composed within and across multiple modalities based on time-cue co-exist with the IoT streaming data. In an audiovisual learning example, because consecutive visual frames usually comprise the same object, this approach provides a unique way to organize features from the same object together. The same method can also organize visual object features with the object's spoken-name features together if the spoken name is presented with the object at about the same time. This cross-modality feature organization may further help the organization of visual features that belong to similar objects but acquired at different location and time. Promising results are achieved through evaluations.

Motivated by the existing difficulties in establishing mathematical models and in observing the system state time series for some complex systems, especially for those driven by non-Gaussian Levy motion, we devise a method for extracting non-Gaussian governing laws with observations only on mean exit time. It is feasible to observe mean exit time for certain complex systems. With the observations, a sparse regression technique in the least squares sense is utilized to obtain the approximated function expression of mean exit time. Then, we learn the generator and further identify the stochastic differential equations through solving an inverse problem for a nonlocal partial differential equation and minimizing an error objective function. Finally, we verify the efficacy of the proposed method by three examples with the aid of the simulated data from the original systems. Results show that the method can apply to not only the stochastic dynamical systems driven by Gaussian Brownian motion but also those driven by non-Gaussian Levy motion, including those systems with complex rational drift.

A stochastic averaging technique based on energy-dependent frequency is extended to dynamical systems with triple-well potential driven by colored noise. The key procedure is the derivation of energy-dependent frequency according to the four different motion patterns in triple-well potential. Combined with the stochastic averaging of energy envelope, the analytical stationary probability density (SPD) of tri-stable systems can be obtained. Two cases of strongly nonlinear triple-well potential systems are presented to explore the effects of colored noise and validate the effectiveness of the proposed method. Results show that the proposed method is well verified by numerical simulations, and has significant advantages, such as high accuracy, small limitation and easy application in multi-stable systems, compared with the traditional stochastic averaging method. Colored noise plays a significant constructive role in modulating transition strength, stochastic fluctuation range and symmetry of triple-well potential. While, the additive and multiplicative colored noises display quite different effects on the features of coherence resonance (CR). Choosing a moderate additive noise intensity can induce CR, but the multiplicative colored noise cannot., Comment: 14 pages, 9 figures

Motivated by improving performance of a bi-stable vibration energy harvester (VEH) from the viewpoint of vibration control, the time-delayed feedback control of displacement and velocity are constructively proposed into an electromechanical coupled VEH mounted on a rotational automobile tire, which is subject to colored noise and the periodic excitation. Using the improved stochastic averaging procedure based on energy-dependent frequency, the expressions of stationary probability density (SPD) and signal-to-noise ratio (SNR) are obtained analytically. Then, the efficiency of time-delayed feedback control on the stationary response and stochastic resonance (SR) for the delay-controlled VEH is explored in detail theoretically. Results show that both noise-induced SR and delay-induced SR can occur. Time delay is able to not only enhance the SR behavior but also weaken it. Furthermore, a larger negative feedback gain of displacement and a larger positive feedback gain of velocity are more beneficial for VEH. Interesting finding is that the optimal combination of time delay in maximizing the harvested performance, such as the harvest power, the output RMS voltage and the power conversion efficiency, is almost perfectly consistent with that in maximizing SNR. Compared with the uncontrolled VEH, the delay-controlled VEH can achieve certain desirable optimization in harvesting energy by choosing the appropriate combination of time delays and feedback gains., Comment: 16 pages, 6 figures

The emergence of the exit events from a bounded domain containing a stable fixed point induced by non-Gaussian L\'evy fluctuations plays a pivotal role in practical physical systems. In the limit of weak noise, we develop a Hamiltonian formalism under the L\'evy fluctuations with exponentially light jumps for one- and two-dimensional stochastic dynamical systems. This formalism is based on a recently proved large deviation principle for dynamical systems under non-Gaussian L\'evy perturbations. We demonstrate how to compute the most probable exit path and the quasi-potential by several examples. Meanwhile, we explore the impacts of the jump measure on the quasi-potential quantitatively and on the most probable exit path qualitatively. Results show that the quasi-potential can be well estimated by an approximate analytical expression. Moreover, we discover that although the most probable exit paths are analogous to the Gaussian case for the isotropic noise, the anisotropic noise leads to significant changes of the structure of the exit paths. These findings shed light on the underlying qualitative mechanism and quantitative feature of the exit phenomenon induced by non-Gaussian noise., Comment: 20 pages, 9 figures