| Safe Haskell | None |
|---|---|
| Language | Haskell2010 |
DataFrame.Boosting.GBM
Description
Gradient boosting of regression trees (Friedman). Trees are fitted to the
negative gradient of the loss each round and accumulated with a shrinkage
factor; squared error gives regression, logistic deviance gives binary
classification. predict is the additive score; gbProbaExpr /
gbDecisionExpr give the classification probability / decision.
Synopsis
- module DataFrame.Model
- data GBLoss
- data GBConfig = GBConfig {
- gbLoss :: !GBLoss
- gbNEstimators :: !Int
- gbLearningRate :: !Double
- gbMaxDepth :: !Int
- gbSeed :: !Int
- defaultGBConfig :: GBConfig
- data GBModel = GBModel {}
- gbExprAtStage :: Int -> GBModel -> Maybe (Expr Double)
- gbProbaExpr :: GBModel -> Expr Double
- gbDecisionExpr :: GBModel -> Expr Bool
Documentation
module DataFrame.Model
The boosting loss.
Constructors
| SquaredError | |
| LogisticDeviance |
Constructors
| GBConfig | |
Fields
| |
Instances
A fitted gradient-boosting model. gbInit is the constant initial score
(mean, or log-odds for classification); gbTrees are the staged regression
trees.
Constructors
| GBModel | |
Instances
| Show GBModel Source # | |||||
| Predict GBModel Source # | |||||
Defined in DataFrame.Boosting.GBM Associated Types
| |||||
| type Prediction GBModel Source # | |||||
Defined in DataFrame.Boosting.GBM | |||||
gbExprAtStage :: Int -> GBModel -> Maybe (Expr Double) Source #
The prediction expression using only the first k trees (staged predict).