arXiv subjects

Yi Feng

Publications and source records attributed to Yi Feng.

149 records · Page 9Linked to original sources

arXiv2018

The optical/UV excess of X-ray-dim isolated neutron star II. nonuniformity of plasma on strangeon star surface

X-ray-dim isolated neutron stars (XDINSs), also known as the Magnificent Seven, exhibits a Planck-like soft X-ray spectrum. In the optical/ultraviolet(UV) band, there is an excess of radiation compared to the extrapolation from the X-ray spectrum. A model of bremsstrahlung emission from a nonuniform plasma atmosphere is proposed in the regime of a strangeon star to explain the optical/UV excess and its spectral deviation. The nonuniform atmosphere is formed by the accretion of interstellar matter which moves along the magnetic field lines to polar caps, and these particles may spread out of the pole regions. The modeled electron temperatures are $\sim100-200$ eV with radiation radii $R_{\rm opt}^{\infty}\sim5-14$ km. The spectra of five sources (RX J0720.4--3125, RX J0806.4--4123, RX J1308.6+2127, RX J1605.3+3249, RX J1856.5--3754) from optical/UV to X-ray bands could be well fitted by the radiative model, and exhibit gaussian absorption lines at $\sim 100-500$ eV as would be expected. Furthermore, the surroundings (i.e., fallback disks or dusty belts) of XDINSs could be tested by future observations with the Large Optical Telescope (LOT).

astro-ph.HE↗

arXiv2018

Effective Neural Solution for Multi-Criteria Word Segmentation

We present a simple yet elegant solution to train a single joint model on multi-criteria corpora for Chinese Word Segmentation (CWS). Our novel design requires no private layers in model architecture, instead, introduces two artificial tokens at the beginning and ending of input sentence to specify the required target criteria. The rest of the model including Long Short-Term Memory (LSTM) layer and Conditional Random Fields (CRFs) layer remains unchanged and is shared across all datasets, keeping the size of parameter collection minimal and constant. On Bakeoff 2005 and Bakeoff 2008 datasets, our innovative design has surpassed both single-criterion and multi-criteria state-of-the-art learning results. To the best knowledge, our design is the first one that has achieved the latest high performance on such large scale datasets. Source codes and corpora of this paper are available on GitHub.

cs.CL↗

arXiv2018

Dual Long Short-Term Memory Networks for Sub-Character Representation Learning

Characters have commonly been regarded as the minimal processing unit in Natural Language Processing (NLP). But many non-latin languages have hieroglyphic writing systems, involving a big alphabet with thousands or millions of characters. Each character is composed of even smaller parts, which are often ignored by the previous work. In this paper, we propose a novel architecture employing two stacked Long Short-Term Memory Networks (LSTMs) to learn sub-character level representation and capture deeper level of semantic meanings. To build a concrete study and substantiate the efficiency of our neural architecture, we take Chinese Word Segmentation as a research case example. Among those languages, Chinese is a typical case, for which every character contains several components called radicals. Our networks employ a shared radical level embedding to solve both Simplified and Traditional Chinese Word Segmentation, without extra Traditional to Simplified Chinese conversion, in such a highly end-to-end way the word segmentation can be significantly simplified compared to the previous work. Radical level embeddings can also capture deeper semantic meaning below character level and improve the system performance of learning. By tying radical and character embeddings together, the parameter count is reduced whereas semantic knowledge is shared and transferred between two levels, boosting the performance largely. On 3 out of 4 Bakeoff 2005 datasets, our method surpassed state-of-the-art results by up to 0.4%. Our results are reproducible, source codes and corpora are available on GitHub.

cs.CL↗

arXiv2014

Early turbulent mixing as the origin of chemical homogeneity in open star clusters

The abundances of elements in stars are a critical clue to their origins. Observed star-to-star variations in logarithmic abundance within an open cluster are typically only $\sim 0.01-0.05$ over many elements, significantly smaller than the variation of $\sim 0.06-0.3$ seen in the interstellar medium from which the stars form. It is unknown why clusters are so homogenous, and whether homogeneity should also prevail in regions of lower star formation efficiency that do not produce bound clusters. Here we report adaptive mesh simulations using passively-advected scalars in order to trace the mixing of chemical elements as star-forming clouds form and collapse. We show that turbulent mixing during cloud assembly naturally produces a stellar abundance scatter at least ~5 times smaller than that in the gas, sufficient to fully explain the observed chemical homogeneity of stars. Moreover, mixing occurs very early, so that regions with efficiencies $\varepsilon \sim 10\%$ are nearly as well-mixed as those with $\varepsilon\sim 50\%$. This implies that even regions that do not form bound clusters are likely to be well-mixed, and enhances the prospects for using chemical tagging to reconstruct dissolved star clusters via their unique chemical signatures.

astro-ph.GA↗

arXiv2012

Pattern Detection with Rare Item-set Mining

The discovery of new and interesting patterns in large datasets, known as data mining, draws more and more interest as the quantities of available data are exploding. Data mining techniques may be applied to different domains and fields such as computer science, health sector, insurances, homeland security, banking and finance, etc. In this paper we are interested by the discovery of a specific category of patterns, known as rare and non-present patterns. We present a novel approach towards the discovery of non-present patterns using rare item-set mining.

cs.SE↗