Last Updated: 2011aug22

CHASC/C-BAS

Topics in Astrostatistics

Statistics 310, Harvard University
Statistics 281, University of California, Irvine

Fall/Winter/Spring 2010-2011

[this course is not being offered at Harvard this year]
Schedule Tuesdays 3:30PM - 5:30PM EST
Location CfA 60 Garden P-226 (Tea Room)



Presentations
Group
13 Sep 2010
Noon-2pm
Projects and timetable for the year.
 
Astrostat Haiku
27 Sep 2010
Noon-2pm
Alanna Connors : Quantifying Doubt and Confidence in Image "Deconvolution"
[.pptx] ; [.fits]
Frank Primini : Computing Average Source Intensity for X-ray Sources Observed in Multiple X-ray Images
[.pdf]
Jennifer Posson-Brown : Power-laws and Solar Flares
[.pdf] ; [.mov]
Jaesub Hong : Looking for X-ray Modulation without relying on X-ray Modulation
[.ppt]
Kaisey Mandel : Hierarchical Bayesian Models for Type Ia Supernova Light Curves, Dust, and Cosmic Distances
[.pdf]
Pavlos Protopapas : Time Series Fitting
[.ppt]
 
Pavlos Protopapas
18 Oct 2010
Fitting Time Series Parametrically
Abstract: Determining periodic or quasi periodic signals, finding low signal-to-noise events and determining its nature in large datasets of astronomical time series is the focus of this talk. I will describe few astronomical transient phenomena as well as current and upcoming new surveys that will exasparate the need of new methodologies to identifying and fitting the nature of the signals. These methods need to leverage statistical and machine learning techniques to cope with the low signal-to-noise ration and the stupendous size of the datasets. The rest of the talk will be dedicated to current efforts and open questions.
[.ppt] ; [.mov]
 
Andreas Zezas (Crete) & Vinay Kashyap (CfA)
09 Nov 2010
11:30 am
logN-logS
Abstract: We will discuss the use in Astronomy of cumulative number distributions, typified as logN-logS curves. These curves are useful to determine large scale information, and even serve as useful constraints on cosmology. We will elaborate on specific cases, as well as discuss biases and observational complications. Finally, we will point to ways in which a hierarchical analysis of the same datasets could be used to extract luminosity functions.
VK slides [.pdf]
AZ slides [.ppt]
 
Aneta Siemiginowska (CfA)
16 Dec 2010
Pratt
12:30pm EST
Investigating gamma-ray properties of young compact radio sources with Fermi.
Abstract: Theoretical models predict that a significant fraction of the energy of a young radio source should be radiated in gamma-rays. However, these sources are distant and their gamma-ray emission is weak. Therefore they could not have been detected by gamma-ray observations before Fermi Gamma-ray Space Telescope. The Fermi sensitivity reaches detection limits of many of these sources, but there have been no reported Fermi detection of a young radio source to date. In our project we explore available Fermi/LAT observations and study observational properties of these objects and will carry out a series of investigations of the data. I will describe a concept of a full statistical model based on a Bayesian approach to evaluate the gamma-ray flux distribution of young radio sources. I also present some potential issues with this model and open questions about our approach.
 
Kaisey Mandel (CfA)
25 Jan 2011
3:30pm EST
Hierarchical Bayesian Models for Type Ia Supernova Light Curves, Dust, and Cosmic Distances
Abstract: Type Ia supernovae (SN Ia) are the most precise cosmological distance indicators and are important for measuring the acceleration of the Universe and the properties of dark energy. To obtain the best distance estimates, the photometric time series (apparent light curves) of SN Ia at multiple wavelengths must be properly modeled. The observed data result from multiple random and uncertain effects, such as measurement error, host galaxy dust extinction and reddening, peculiar velocities, and distances. Furthermore, the intrinsic, absolute light curves of SN Ia differ between individual events: different SN Ia have different intrinsic luminosities, colors and light curve shapes, and these properties are correlated in the population. A hierarchical Bayesian model provides a natural statistical framework for coherently accounting for these multiple random effects while fitting individual SN Ia and the population distribution. I will discuss the application of this statistical model to optical and near-infrared data for computing inferences about the dust, distances and intrinsic covariance structure of SN Ia. Using this model, I demonstrate that the combination of optical and NIR data improves the precision of SN Ia distance predictions by about a factor of 2 compared to using optical data alone. Finally, I will discuss some open research problems concerning statistical analysis of supernova data and their application to cosmology.
[astro-ph:1011.5910]
Presentation slides [.pdf]
 
Nathan Stein (Harvard)
8 Feb 2011
3:30pm EST
Segregating solar features by temperature
Abstract: To investigate the thermal properties of the solar corona, images of the Sun are observed in multiple wavelengths. Efficiently combining the information from multiple images into a temperature map of the Sun is necessary in order to take full advantage of the enormous amount of data arriving as high spatial and temporal resolution images. I will discuss a clustering method for identifying regions of the Sun with similar thermal properties, while avoiding the computational expense of reconstructing the temperature distribution in each pixel.
Presentation slides [.pdf]
 
Ashish Mahabal (CalTech)
15 Feb 2011
Aspects of Transient Classification
Abstract: Various ongoing and forthcoming synoptic surveys provide a unique opportunity to explore the variable sky. An ever growing number of transients will be detected per night. A majority of these belong to fairly well understoodclasses on which one need not waste the scarce follow-up resources. As a result selecting which transients to follow becomes more and more critical for understanding newer and/or rarer classes. The inputs are diverse and not easy to make a cohesive sense of especially when one is interested in the classification in as close to real-time as possible. We will present various aspects of this process, the current methodologies and understanding, and a sense of where we are heading.
Presentation slides [.pdf]
 
Jae Sub Hong (CfA)
29 Mar 2011
Noon
Looking for true modulation period through energy quantiles
Abstract: We have discovered 10 periodic X-ray sources and 11 candidates in the 1 Ms chandra ACIS observation of the Limiting Window, a low extinction region at 1.4 deg south of the Galactic center. Their X-ray and optical properties are consistent with those for magnetic cataclysmic variables (MCVs), which presumably account for a large number (>3000) of the low luminosity hard X-ray sources in the Bulge. The sheer number of the Bulge X-ray sources found in the Galactic center region indicates the importance of their nature in understanding the formation and evolutionary history of the inner Galaxy. We used three search routines for finding X-ray periodicity: Lomb-Scargle routine, Buccheri's z2 statistics, the Epoch Folding method. Multiple periods are found in some sources, and it is often difficult to separate the true period from its simple harmonics. We use energy quantiles in attempt to validate or find the periodicity that truly represents the emission geometry.
Presentation slides: [.pptx] ; [.pdf]
 
David Stenning (UC Irvine)
29 Mar 2011
1 pm
Automatic Classification of Sunspot Groups Using SOHO/MDI Magnetogram and White-Light Images
Abstract: While solar data is being generated at an unprecedented rate, most sunspot classification is still done manually by experts. This is a labor-intensive process and, as with all manual procedures, is susceptible to human observer biases. Using SOHO/MDI magnetogram and white-light images, we propose a system for automatically classifying sunspot groups into four broad classes based on the Mount Wilson scheme. This scheme uses magnetic active-region structure as seen in magnetogram images, specifically areas of opposite polarity magnetic flux, to classify sunspot groups using simple rules. By utilizing techniques from mathematical morphology, we extract features that can be used in classification algorithms to produce an automatic sunspot classification procedure. I will discuss the progress we have made as well as plans for future work.
Presentation slides [.pdf]
 
Antonaldo Diaferio (Universita' degli Studi di Torino)
19 Apr 2011
Noon
SciCen 705
The expansion history of the Universe: myths and facts
Abstract: Data from extensive surveys of high-redshift type Ia supernovae and gamma-ray bursts have usually been interpreted as a robust indication that the Universe expansion had an initial deceleration phase followed by the present acceleration phase. When analysed with a proper Bayesian approach, this interpretation is not in fact as robust as usually assumed. I show that the expansion history predicted by conformal gravity, which is substantially different from the standard model, can accommodate the data equally well. The Bayesian evidence is required to discriminate between the two models.
Diaferio, Ostorero, & Cardone, Gamma-ray bursts as cosmological probes: ΛCDM vs. conformal gravity, 2011, arXiv:1103.5501
Presentation: [.pdf]
 
Jin Xu (UC Irvine)
3 May 2011
pyBLoCXS
Abstract:
Slides [.pdf]
 
Brandon Kelly (CfA)
31 May 2011
Noon
SciCen 705 706
Modeling active galactic nuclei variability with stochastic processes
Abstract:
I describe a statistical model for the X-ray fluctuations of accreting black holes. The model is formulated in the time domain via a set of stochastic differential equations, and I describe a Bayesian approach for performing statistical inference using the model. Specifically, I model the X-ray fluctuations as a mixture of Ornstein-Uhlenbeck processes with varying relaxation time scales. The mixture of OU-processes is derived as the solution to the stochastic diffusion equation, enabling an astrophysical interpretation of the results. The technique is not biased by red noise leak, aliasing, irregular sampling, and measurement error, and is computationally efficient. We apply our model to the X-ray time series of 10 local AGN and show that our model is both a good fit to the data, and is able to recover previous results with increased precision. We recover the previously known correlation between the black hole mass and characteristic time scale of the X-ray fluctuations, and find a tight anti-correlation between the black hole mass and the amplitude of the driving noise field in our model, which is proportional to the amplitude of the high frequency X-ray PSD.
Slides [.ppt]
 
Paul Baines and Irina Udaltsova (UCDavis)
12 Jul 2011
9am PDT/Noon EDT
SciCen 706
Bayesian estimation of logN-logS
Abstract: The study of source populations is often conducted using the cumulative distribution of the number of sources detected at a given sensitivity. The resulting "log(N>S)-logS" distribution can be used to compare and evaluate theoretical models for source populations and their evolution. In practice, however, inferring properties of source populations from observational data is complicated by the presence of detector-induced uncertainty and bias. This includes background contamination, uncertainty on both intensity and location of sources, and, most challenging, the issue of non-detections or unobserved sources. Since the probability of a non-detection is a function of the unobserved flux, the missing data mechanism is non-ignorable. We present a computationally efficient Bayesian approach for inferring physical model parameters and the corrected log(N>S)-log(S) distribution for source populations. Our method extends existing work in allowing for both non-ignorable missing data and an unknown number of unobserved sources. Importantly, our method is also scalable in the number of observed sources, and computationally insensitive to the number of missing sources. By correcting for the non-ignorable missing data mechanism and other detection phenomena, we are able to obtain corrected estimates of the flux and luminosity distribution of source populations.
SCMA V poster [.pdf]
Presentation slides (updated) [.pdf]
 
Mark Weber (SAO)
26 Jul 2011
Noon EDT
SciCen 706
Characterizing Underconstrained DEM Analysis
Abstract: Differential emission measures (DEMs) are one of the principal ways that solar astronomers derive physical properties of the optically thin corona from observations. Imaging instruments (e.g., the Atmospheric Imaging Assembly on the Solar Dynamics Observatory) have several advantages over spectroscopic instruments, such as cadence, field of view, and spatial resolution, but do not have enough independent channels to adequately constrain the temperature distributions of the observed plasmas. Thus, DEM analysis with these data sets is an ill-posed, underconstrained linear algebra problem. I discuss some of the shortcomings of solution techniques in the literature; in particular, none of the extant methods provide a way to distinguish and identify the infinite set of solutions to the underconstrained problem as a subset of all possible DEMs, which has consequences for physical interpretations. I present the Convex-Hull method of DEM analysis, which overcomes several of the pitfalls of DEM analysis, and which also highlights an interesting aspect of the instrument response functions that has implications for setting up a Bayesian framework for error propagation. I have not yet set up such a framework and look forward to some interesting discussion on that point.
Presentation [.pptx]
 
 
 
Fall/Winter 2004-2005
Siemiginowska, A. / Connors, A. / Kashyap, V. / Zezas, A. / Devor, J. / Drake, J. / Kolaczyk, E. / Izem, R. / Kang, H. / Yu, Y. / van Dyk, D.
Fall/Winter 2005-2006
van Dyk, D. / Ratner, M. / Jin, J. / Park, T. / CCW / Zezas, A. / Hong, J. / Siemiginowska, A. & Kashyap, V. / Meng, X.-L.
Fall/Winter 2006-2007
Lee, H. / Connors, A. / Protopapas, P. / McDowell, J., / Izem, R. / Blondin, S. / Lee, H. / Zezas, A., & Lee, H. / Liu, J.C. / van Dyk, D. / Rice, J.
Fall/Winter 2007-2008
Connors, A., & Protopapas, P. / Steiner, J. / Baines, P. / Zezas, A. / Aldcroft, T.
Fall/Winter 2008-2009
H. Lee / A. Connors, B. Kelly, & P. Protopapas / P. Baines / A. Blocker / J. Hong / H. Chernoff / Z. Li / L. Zhu (Feb) / A. Connors (Pt.1) / A. Connors (Pt.2) / L. Zhu (Mar) / E. Kolaczyk / V. Liublinska / N. Stein
Fall/Winter 2009-2010
A.Connors / B.Kelly / N.Stein, P.Baines / D.Stenning / J. Xu / A.Blocker / P.Baines, Y.Yu / V.Liublinska, J.Xu, J.Liu / X.L. Meng, et al. / A. Blocker, et al. / A. Siemiginowska / D. Richard / A. Blocker / X. Xie / X. Jin / V. Liublinska / L. Jing
AcadYr 2010-2011
Astrostat Haiku / P. Protopapas / A. Zezas & V. Kashyap / A. Siemiginowska / K. Mandel / N. Stein / A. Mahabal / J.S. Hong / D. Stenning / A. Diaferio / X. Jin / B. Kelly / P. Baines & I. Udaltsova / M. Weber
AcadYr 2011-2012

CHASC