columna API reference - v0.5.0
    Preparing search index...

    Function treeNet

    • TreeNet-style gradient boosting.

      • regression: squared-error residual trees
      • classification: multinomial deviance (K=2 → Bernoulli) with one tree per class per stage

      Parameters

      • X: ArrayLike<ArrayLike<number>>
      • y: ArrayLike<string | number>
      • Optionaloptions: {
            learningRate?: number;
            maxDepth?: number;
            nTrees?: number;
            seed?: number;
            task?: "regression" | "classification";
        }

      Returns TreeNetResult

    Settings
    Member Visibility