| 1. |
|
|
| 2. |
- Galan, C., et al.
(författare)
-
International observational campaigns of the last two eclipses in EE Cephei : 2003 and 2008/9
- 2012
-
Ingår i: Astronomy and Astrophysics. - : EDP Sciences. - 0004-6361 .- 1432-0746. ; 544, s. A53-
-
Tidskriftsartikel (refereegranskat)abstract
- Context. EECep is an unusual long-period (5.6 yr) eclipsing binary discovered during the mid-twentieth century. It undergoes almost-grey eclipses that vary in terms of both depth and duration at different epochs. The system consists of a Be type star and a dark dusty disk around an invisible companion. EECep together with the widely studied epsilon Aur are the only two known cases of long-period eclipsing binaries with a dark, dusty disk component responsible for periodic obscurations.Aims. Two observational campaigns were carried out during the eclipses of EECep in 2003 and 2008/9 to verify whether the eclipsing body in the system is indeed a dark disk and to understand the observed changes in the depths and durations of the eclipses.Methods. Multicolour photometric data and spectroscopic observations performed at both low and high resolutions were collected with several dozen instruments located in Europe and North America. We numerically modelled the variations in brightness and colour during the eclipses. We tested models with different disk structure, taking into consideration the inhomogeneous surface brightness of the Be star. We considered the possibility of disk precession.Results. The complete set of observational data collected during the last three eclipses are made available to the astronomical community. The 2003 and 2008/9 eclipses of EECep were very shallow. The latter is the shallowest among all observed. The very high quality photometric data illustrate in detail the colour evolution during the eclipses for the first time. Two blue maxima in the colour indices were detected during these two eclipses, one before and one after the photometric minimum. The first (stronger) blue maximum is simultaneous with a "bump" that is very clear in all the UBV(RI)(C) light curves. A temporary increase in the I-band brightness at the orbital phase similar to 0.2 was observed after each of the last three eclipses. Variations in the spectral line profiles seem to be recurrent during each cycle. The Na I lines always show at least three absorption components during the eclipse minimum and strong absorption is superimposed on the H alpha emission.Conclusions. These observations confirm that the eclipsing object in EECep system is indeed a dark, dusty disk around a low luminosity object. The primary appears to be a rapidly rotating Be star that is strongly darkened at the equator and brightened at the poles. Some of the conclusions of this work require verification in future studies: (i) a complex, possibly multi-ring structure of the disk in EECep; (ii) our explanation of the "bump" observed during the last two eclipses in terms of the different times of obscuration of the hot polar regions of the Be star by the disk; and (iii) our suggested period of the disk precession (similar to 11-12 P-orb) and predicted depth of about 2(m) for the forthcoming eclipse in 2014.
|
|
| 3. |
|
|
| 4. |
- Oberst, Michael, et al.
(författare)
-
Characterization of Overlap in Observational Studies
- 2020
-
Ingår i: Proceedings of Machine Learning Research. - 2640-3498. ; 108, s. 788-797
-
Konferensbidrag (refereegranskat)abstract
- Overlap between treatment groups is required for non-parametric estimation of causal effects. If a subgroup of subjects always receives the same intervention, we cannot estimate the effect of intervention changes on that subgroup without further assumptions. When overlap does not hold globally, characterizing local regions of overlap can inform the relevance of causal conclusions for new subjects, and can help guide additional data collection. To have impact, these descriptions must be interpretable for downstream users who are not machine learning experts, such as policy makers. We formalize overlap estimation as a problem of finding minimum volume sets subject to coverage constraints and reduce this problem to binary classification with Boolean rule classifiers. We then generalize this method to estimate overlap in off-policy policy evaluation. In several real-world applications, we demonstrate that these rules have comparable accuracy to black-box estimators and provide intuitive and informative explanations that can inform policy making.
|
|
