The virgin non-parametric regression THE CHAD LINEAR REGRESSION With 20 data points, ends up Can be made asymptotically Computationally intensive, being published in top efficient just in a couple of

The virgin non-parametric regression THE CHAD LINEAR REGRESSION
With 20 data points, ends up Can be made asymptotically
Computationally intensive, being published in top efficient just in a couple of steps
Requires a tonne of observations makes 2019 computers bleed macroeconomic journals
for convergence Gives a nice, clean
Works in one step i ior
Does not have a clear interpretation any
Heavily depends on notion of efficient estimation without fine-tuning grandma can comprehend
hyper-parameters that
must be chosen optimally Can be computed Constant partial effect
Even after estimation, the entire support
partial effects require Я
Choice of hyper-parameters computation
needs orders of magnitude
more computational power 1, 3
Is rarely used in applied
economic analysis
Inference can be
closed-form solution
impossible to find simple to the one
for inference algebra libraries
inefficient, highlevel
implementations are slower Most courses spend Can handle hundreds
than Crysis on Pentium hours just to get to of regressors with ease
Parzen—Rozenblatt,
о and then sto ackage that can
curse of Lots of
Implementations vary
to
Conditions yieldi widely in software, Could not care fess about к.
normality error distribution, beginners
estimator are scaling being normal anyway Gives the same
hard to verify factors к consistent result in
Fails miserably in regions No one bats an eye В, STATA, SAS, SPSS.
of low density of Behaves like a spoiled when parametric of explanatory
variables brat at boundaries, variables does not matter
Convergence so poor, has to be Can account for arbitrary
with parametric assumptions or index restrictions including
junctions of regressors,
Even bootstrapped confidence intervals are not trusted at first glance No questions asked with robust standard