Experts: Making Expert Objects

In rai: Revisiting-Alpha-Investing for Polynomial Regression

Description

Experts are the "actors" which "bid" to see who conducts the next test. They contain an object "bidder" that determines bidding strategy and an object "constructor" that determines which feature it wants to text next. The runAuction function calls functions from experts and gWealth. The makeExpert function is not called directly, but through makeStepwiseExpert or makeScavengerExpert. Defaults are not set because these are internal functions called by rai and runAuction and all arguments are required.

Usage

makeExpert(bidder, constructor)

makeStepwiseExpert(gWealth, ncolumns)

makeScavengerExpert(gWealth, theModelFeatures, name)

Arguments

bidder	bidder object; output of makeStepwiseBidder.
constructor	constructor object; output of makeRawSource or makeLocalScavenger.
gWealth	global wealth object, output of gWealthStep.
ncolumns	number of features the constructor should manage, thought of as columns of the design matrix.
theModelFeatures	list of feature names in the model when the feature was rejected.
name	name of base feature used in interactions with other features in the model.

Value

A closure containing a list of functions.

rai documentation built on July 2, 2019, 5:02 p.m.