Sloppy is the term used to describe a class of complex models exhibiting large parameter uncertainty when fit to data.
the Fisher information matrix (FIM) can be used to estimate the uncer- tainty in each parameter in our model.
Sloppiness and Parameter Identifiability, Information Geometry by Mark Transtrum
Input distribution lies in a manifold when there is structure/correlations in high dimensional space
Sloppy Models – Why is science possible?
We describe the possible system behaviors as points in a 'behavior' space, and we find that they form a hyper-ribbon, long along stiff combinations and very thin (reflecting unchanging behavior) when sloppy combinations are changed (fig below right).
"Many models in biology, engineering and physics have a very large number of parameters. Often many of these are only known approximately. Moreover, in John von Neuman's famous quip \with four parameters I can fit an elephant, and with five I can make him wiggle his trunk." suggests that only a small set of these parameters are actually relevant? Could there be a fundamental theory of these Complex systems that allows us to work out what the key parameters are?"
Perspective: Sloppiness and emergent theories in physics, biology, and beyond publication
Parameter Space Compression Underlies Emergent Theories and Predictive Models
Universally Sloppy Parameter Sensitivities in Systems Biology Models
Sloppy-model universality class and the Vandermonde matrix