Results for "local approximation"

Fast approximation of costly simulations.

Techniques to understand model decisions (global or local), important in high-stakes and regulated settings.

Local surrogate explanation method approximating model behavior near a specific input.

Optimization problems where any local minimum is global.

Optimization with multiple local minima/saddle points; typical in neural networks.

Matrix of second derivatives describing local curvature of loss.

Neural networks can approximate any continuous function under certain conditions.

Optimization under uncertainty.

Minimum relative to nearby points.