A5101542729- Stellar, planetary, and galactic studies
- Astronomy and Astrophysical Research
- Astronomical Observations and Instrumentation
- E-commerce and Technology Innovations
- Safety and Risk Management
- Research studies in Vietnam
- Accounting and Organizational Management
- Electron and X-Ray Spectroscopy Techniques
- Big Data and Business Intelligence
- ICT Impact and Policies
- Industrial Vision Systems and Defect Detection
- Terahertz technology and applications
- Non-Destructive Testing Techniques
- Astrophysics and Star Formation Studies
- Franchising Strategies and Performance
- Particle Accelerators and Free-Electron Lasers
- Higher Education and Teaching Methods
- Welding Techniques and Residual Stresses
- Risk Management in Financial Firms
- Financial Distress and Bankruptcy Prediction
- Customer churn and segmentation
- Evaluation and Optimization Models
- Gamma-ray bursts and supernovae
- Gyrotron and Vacuum Electronics Research
- Advanced Electron Microscopy Techniques and Applications
Sichuan Tourism University
2023
Chinese Academy of Sciences
2009-2022
University of Chinese Academy of Sciences
2009-2022
National Astronomical Observatories
2021-2022
Hunan University
2019
Jiangxi Science and Technology Normal University
2012-2014
Shanghai Institute of Applied Physics
2009-2010
Shaoguan University
2008
We developed a convolutional neural network (CNN) model to distinguish the double-lined spectroscopic binaries (SB2s) from others based on single exposure medium-resolution spectra ($R\sim 7,500$). The training set consists of large mock stars and synthesized MIST stellar evolutionary ATLAS9 atmospheric model. Our reaches novel theoretic false positive rate by adding proper penalty negative sample (e.g., 0.12\% 0.16\% for blue/red arm when parameter $\Lambda=16$). Tests show that performance...
Understanding human preferences is crucial for improving foundation models and building personalized AI systems. However, are inherently diverse complex, making it difficult traditional reward to capture their full range. While fine-grained preference data can help, collecting expensive hard scale. In this paper, we introduce Decomposed Reward Models (DRMs), a novel approach that extracts from binary comparisons without requiring annotations. Our key insight represent as vectors analyze them...
Massive binaries play significant roles in many fields. Identification of massive stars, particularly binaries, is great importance. In this paper, by adopting the technique measuring equivalent widths several spectral lines, we identified 9,382 early-type stars from LAMOST medium-resolution survey and divided sample into four groups, T1 ($\sim$O-B4), T2 ($\sim$B5), T3 ($\sim$B7), T4 ($\sim$B8-A). The relative radial velocities $RV_{\rm rel}$ were calculated using Maximum Likelihood...
Binary stars plays important role in the evolution of stellar populations . The intrinsic binary fraction ($f_{bin}$) O and B-type (OB) LAMOST DR5 was investigated this work. We employed a cross-correlation approach to estimate relative radial velocities for each spectra. algorithm described by \cite{2013A&A...550A.107S} implemented several simulations were made assess performance approach. OB are estimated through comparing uni-distribution between observations with Kolmogorov-Smirnov...
A 30-MeV femto-second electron linac is built at the Shanghai Institute of Applied Physics, which can produce high power, coherent THz undulator radiation. We report experimental facility and measurement frequency spectrum. First experiments show averaged power to be about 20 mW.
With the rapid development of e-commerce, more and companies have to carry out precise marketing in order win competition. This article introduces information technologies involved precision marketing.
Modern colleges and universities are basically facing a prominent problem, how to combine teaching, scientific research resources cultivate high-quality talents. In order solve this phenomenon, paper takes university in Sichuan as the object, discusses rational combination of teaching from small paper. At same time, puts forward model 'demand assistance-collaboration drive-collaboration sharing', which focuses on cultivating students' ability master professional knowledge problems field...
Based on the femtosecond accelerator device which was built at Shanghai Institute of Applied Physics (SINAP), recently a modified far infrared Michelson interferometer has been developed to measure length electron bunches via optical autocorrelation method. Compared with our former normal interferometer, we use hollow retroreflector instead flat mirror as reflective mirror. The experimental setup and results bunch measurement will be described in this paper.
Abstract Convolutional neural network (CNN) is an efficient and robust method which can accurately detect the Tailor Rolled Blank laser welding pool penetration status. To select proper hyperparameters optimization of CNN model are black box problem. In this paper, innovative based on to identify status weld during was introduced. A coaxial monitoring platform set up, as well two-class, three-class four-class datasets created for training validating CNN. The Bayesian Optimization (BO) used...
K-means algorithm has powerful ability to cluster large data sets due its high efficiency in mining but calculation instability limits the application of algorithm, so research intelligent optimization become a hot field for researchers related. First original is analyzed with more details; Second, improvement seed selection methods and flow are redesigned speed up enhance stability improved model; Third, paper realizes conducts analysis customer classification practice which show that...
BP neural network is a hot research field for its powerful simulation calculation ability in various disciplines recent years, but the algorithm has some shortages such as low convergence which limit usage of algorithm. The paper improves model with genetic and applies it to evaluate competitive advantages logistics enterprises. First designs an evaluation indicator system advantage enterprises through analyzing characteristics indicator; Second, used speed up based on this advances new...
Artificial intelligence is the path choice for development of industry and enterprise intelligence. In era intelligence, artificial its supporting data technologies are increasingly becoming driving force development, but these new data-dependent smart platforms enabling models still unclear. Thus further comb analysis evolution path, big can assign rules enterprise, relying on intensive, super calculate power, combined to form a algorithm model China, matching corresponding model,...
The key is to make integration of the supply chain system, lower cost agricultural products in international market, improve competitiveness agriculture and increase income farmers from chain. This paper concept products. characteristics are discussed paper. Then, this analyzes model based on management. Finally, puts forward establishment China's solutions. purpose help develop