[BTeam] sloppiness
Jay Benesch
benesch at jlab.org
Tue Aug 23 07:15:02 EDT 2016
I was unaware that sloppiness was a defined mathematical concept. I
haven't gotten far enough into the paper to learn if there's application
to desks.
http://arxiv.org/abs/1608.05679
The geometry of sloppiness
Emilie Dufresne, Heather A. Harrington, Dhruva V. Raman
(Submitted on 19 Aug 2016)
Mathematical models in the sciences often require the estimation of
unknown parameter values from data. Sloppiness provides information
about the uncertainty of this task. We develop the precise mathematical
foundation for sloppiness and define rigorously its key concepts, such
as `model manifold' in relation to concept of structural
identifiability. The traditional definition of sloppiness uses the
Fisher Information Matrix, and as such it deals with infinitesimal
measurement error. We generalize sloppiness and define it in terms of
the premetric on parameter space induced by measurement noise.
Applications include parametric statistical models, explicit time
dependent models, and ordinary differential equation models with time
series data.
Comments: 25 pages, 5 figures
Subjects: Metric Geometry (math.MG); Dynamical Systems (math.DS);
Statistics Theory (math.ST); Data Analysis, Statistics and Probability
(physics.data-an); Quantitative Methods (q-bio.QM)
More information about the BTeam
mailing list