Consider the following constructors:
public FastTreeBinaryClassificationTrainer(IHostEnvironment env,
string labelColumn,
string featureColumn,
string weightColumn = null,
int numLeaves = Defaults.NumLeaves,
int numTrees = Defaults.NumTrees,
int minDocumentsInLeafs = Defaults.MinDocumentsInLeafs,
double learningRate = Defaults.LearningRates,
Action<Arguments> advancedSettings = null)
public FastTreeRegressionTrainer(IHostEnvironment env,
string labelColumn,
string featureColumn,
string weightColumn = null,
int numLeaves = Defaults.NumLeaves,
int numTrees = Defaults.NumTrees,
int minDocumentsInLeafs = Defaults.MinDocumentsInLeafs,
double learningRate = Defaults.LearningRates,
Action<Arguments> advancedSettings = null)
public FastTreeRankingTrainer(IHostEnvironment env, string labelColumn, string featureColumn, string groupIdColumn,
string weightColumn = null, Action<Arguments> advancedSettings = null)
There is no reason to have disparity here: ranker should expose the same numLeaves, numTrees etc.
The only difference should be the presence of required groupId in the ranker constructor.
Consider the following constructors:
There is no reason to have disparity here: ranker should expose the same
numLeaves,numTreesetc.The only difference should be the presence of required
groupIdin the ranker constructor.