Public Seminar of Research Postgraduate Programmes(RPg) Student:- Application of a Globular Cluster Evolution Model in the Gamma Ray Excess at the Galaxy Centers and the Formation of Nuclear Star Clusters
The galaxy center gamma ray excess in the Milky Way(MW) and the Andromeda(M31) detected by Fermi-LAT has raised a debate over its origin being from the proposed dark matter(DM) annihilation or an unresolved population of millisecond pulsars(MSPs). Resolving this problem shall have great influence on where to hunt DM. We have used a new tested model of globular cluster(GC) formation to simulate the luminosity contribution from MSPs, and found that compared with a previous study with a crude GC formation model, the result
1) fits the MW excess even better , but
2) likewaise substantially insufficient to the M31 excess.
This suggests that the two galaxies are more different than they were believed, that the DM distribution and the composition at the nucleus shows discernable discrepancies. However we cannot rule out the possibility that it might be due to our lesser knowledge about the gamma ray sources in M31 than the MW. Furthermore, in carrying out the work we have developed a complete pipeline of GC evolution and mass deposition, which can serve to study its contribution to the formation and properties of the nuclear star cluster(NSC). Our database stores samples of halo masses from 1e9~1e14 msun with different merging histories. Thus we are able to look at the statistics across a large span. We hope to be able to shed lights on various properties of the NSCs such as the metalicity, mass, radius etc.
Kepler Space Telescope and Transiting Exoplanet Survey Satellite(TESS) is used to detect planet candidates by using a convolutional neural network model. The Q1-Q17(DR24) table is applied as our training and test sets. The model takes two phase folded light curves and some parameters of each transit-like signal and then outputs whether the signal represents a planet candidate(PC) a non-transiting phenomena(NTP) or a false positive(FP). In the current model, 17 features is fed into a dense neural network model, such as transit durations and depth of signals. At this stage, the model achieves AUROC and accuracy of about 97.7%, 95.9% respectively for the test set. The accuracy for the training set can be over 99%, which means that the model can easily overfit the data. The most straightforward way to the problem is to use more data to train the model. Therefore, a plan is to train it with more simulated data later in order to increase the AUROC and accuracy of predictions.
時間：2020.10.30(五) 10:30
頻道：Zoom:291-628-0452
講者：Yuan GAO(香港大學)
語言：英語
免費講座，歡迎參加
此活動由香港大學物理系.主辦。
