forestTooltips top-level property

Implementation

Map forestTooltips = {
  AlgorithmType.conditional: '''

    **Conditional:** The conditional forest is an extension of the traditional
    Random Forest. It is focussed on handling situations where the response
    variable is influenced by specific conditions or covariates.  Conditional
    inference focuses on estimating the conditional distribution of the response
    variable given certain predictors. It is particularly useful in causal
    inference and when dealing with heterogeneous treatment effects.

      ''',
  AlgorithmType.traditional: '''

      **Traditional:** The traditional and original random forest algorithm
      (also called bagging or bootstrap aggregation) resamples the original
      dataset multiple times to build multiple decision trees. Each dataset is a
      sample of both the observations and the variables and can lead to reducing
      overfitting and so improving generalisation. The final ensemble model is
      then the aggregate of the predictions of the individual decision trees.

      ''',
};