Charlotte L. Haley

PhD, PStat, Statistical Analyst

Chicago pic

I am an assistant statistical scientist at Argonne National Laboratory, in the Laboratory for Advanced Numerical Simulation (LANS) group within the Division of Mathematics and Computer Science (MCS).


My research interests lie in statistics and signal processing, in particular spectrum analysis of time series, array processing, and spatio-temporal statistics.

Education

Degrees:

PhD (Queen's) in Statistics
Thesis: Nonparametric and Parametric Methods for Solar Oscillation Spectra
Supervised by David J Thomson

MSc (Queen's) in Statistics
  Thesis: Statistical Analysis of Atrial Fibrillation Electrograms
Supervised by Selim G Akl, Damian P Redfearn, and David J Thomson

Honours Thesis (Dalhousie): On Wilson's Theorem
Supervised by Karl Dilcher

Research

Google Scholar Profile
Research Statement

My research interests lie in frequency domain (spectral) analysis of time series and spatiotemporal processes. Nonparametric spectrum analysis is arguably one of the finest and most elegant exploratory tools for scientific inference on physical processes that vary spatially and temporally. The power spectrum of a stationary time series partitions the variance of the signal into its constituent harmonic components by means of the Fourier transform, revealing oscillatory structures and frequency-domain relationships between input and output of linear systems. Temporal correlations in time series often preclude the application of conventional statistical methods for modeling and analysis, but it is possible to model covariance structure in the frequency domain as individual Fourier components are effectively uncorrelated. Several high impact areas of application include space physics, climate science, and seismology.

My doctoral work at Queen’s University focused on false detection rates and shape of large peaks in estimates of the power spectrum. This work was applied to solar physics data [1]. I also collaborated with the division of cardiology at the Kingston General Hospital, where I (i) provided general statistical consulting advice, mainly receiver operating characteristics, T- and Mann-Whitney U tests, and ANOVA analyses [2,3] and (ii) designed an algorithm designed to rate atrial electrogram activity similar to expert cardiologist opinion, and to determine if experts rate consistently [4].

I then began a postdoctoral appointment at Argonne National Laboratory in spatio-temporal statistics for the Department of Energy. My research focused on (a) the prediction of solar irradiance from multivariate time series using periodic (cyclostationary) autoregressive models for the purpose of determining solar farm energy output [6] (b) bandwidth optimization in spectrum analysis of phasor measurement time series [5] (c) spatial and temporal frequency domain analysis of four-dimensional data sets from fluid dynamics simulation output. I am now continuing my appointment at Argonne as an assistant scientist.

[1] D. J. Thomson and C. L. Haley. “Spacing and shape of peaks in non-parametric spectrum estimates”. Proceedings of the Royal Society of London Series A 470.2167 (2014), p. 20140101.
[2] K. A. Michael, C. L. Haley, A. Baranchuk, C. S. Simpson, H. Abdollah, and D. P. Redfearn. “Failed Anti-Tachycardia Pacing can be used to Differentiate Atrial Arrhythmias from Ventricular Tachycardia in Implantable Cardioverter-Defibrillators”. Europace 17.1 (2015), pp. 78–83.
[3] R. Somani, K. Mohajer, C. L. Haley, C. S. Simpson, H. Abdollah, A. Baranchuk, D. P. Redfearn, and K. A. Michael. “The Periprocedural Use of Dabigatran in Patients Undergoing Left Atrial Ablation for Atrial Fibrillation”. Cardiovasc Ther. 32.5 (2014), pp. 198–201.
[4] C. L. Haley, Gula L. J., R. Miranda, A. Baranchuk, K. A. Michael, C. S. Simpson, H. Abdollah, A. J. West, S. G. Akl, and D. P. Redfearn. “Validation of a novel algorithm for quantification of fractionation in atrial fibrillation electrograms.” Europace 15.3 (2013), pp. 447–452.
[5] C. L. Haley and M. Anitescu. “Optimal bandwidth for multitaper spectrum analysis”. IEEE Signal Processing Letters X (2017), pp. xx–xx. Forthcoming.
[6] C. L. Haley and M. Anitescu. “Cyclostationary models of solar irradiance”. In: Joint Statistical Meetings (JSM) Proceedings. American Statistical Association. Seattle, WA, USA, 2015, pp. 1212–1225.

2016-2017 Talks
  1. C.L. Haley, C. J. Geoga, and Anitescu M. Detection and estimation of oscillatory components in power networks. Joint Statistical Meetings, Baltimore MA, Aug/2017.
  2. C.L. Haley. On Spatiotemporal Spectral Analysis Invited Talk, University of Toronto, Toronto On, Feb/2017.
  3. C.L. Haley, Anitescu, M. Optimal bandwidth for multitaper spectrum estimation. Joint Statistical Meetings, Chicago IL, Aug/2016.

Teaching

Primary Instructor

 

In Fall 2011, I was the primary instructor for MTHE 224 at Queen's, a second year civil engineering course. Course material was divided into three components

  1. Statistics (6.5 weeks)
  2. Differential Equations (6.5 weeks)
  3. Matlab labs and tutorials.

I designed lesson plans, course notes, prepared assessments, lectured, and managed over 100 teaching assistant hours.

Teaching Assistant