Research

Summary

My doctoral research was in the intersection of model-based clustering and incomplete data via multiple imputation.

  • Clustering:  Finding grouping patterns in data, with no prior knowledge of group labels.
  • Model-based clustering:  Finding groups, and describing them with statistical models.
  • Incomplete data: Holes or gaps in a data set result in incomplete records, which are often thrown out before analysis begins.
  • Multiple imputation: Create multiple “guesses” at the missing data values, and obtain a single result for your analysis that takes into account all guesses.

I am always looking for new projects and collaborations!  If you have an idea, please send me an email: larosec@newpaltz.edu


Publications & Research

  • Larose, C., Harel, O., Kordas, K., Dey, D. K.  (2016)  Latent class analysis of incomplete data via an entropy-based criterion.  Statistical Methodology 32, 107 – 121.
  • Belliveau, T., Jette, A. M., Seetharama, S., Axt, J., Rosenblum, D., Larose, D., Houlihan, B., Slavin, M., Larose, C. (2016)  Developing artificial neural network models to predict functioning one year after traumatic spinal cord injury.   Archives of Physical Medicine and Rehabilitation.  In press.
  • Larose, C., Dey, D. K., Harel, O.   The Impact of Incomplete Data on Measures of Uncertainty.  Submitted.
  • Larose, C., Harel, O., Yan, J.  Estimating Regression Coefficient Change in Incomplete Data.  In preparation.
  • Larose, C., Dey, D. K., Harel, O.  Clustering incomplete normal mixture data without pre-specification of the number of clusters.  In preparation.