daisugi

Getting Started with daisugi daisugi is an experimental machine learning package focused on unconventional tree systems, probabilistic boosting, interpretable ensembles, and emerging recursive partitioning methods. Most workflows in daisugi follow a simple pattern: “grow_*()” fits a machine “harvest_*()” generates predictions The package primarily uses explicit “(x, y)” interfaces for classification and regression tasks.

Workflow Overview

A typical daisugi workflow:

1. prepare predictors and targets
2. grow a machine
3. harvest predictions
4. evaluate externally using your preferred tooling

daisugi intentionally avoids imposing a modeling framework and instead focuses on exposing novel algorithms through a lightweight interface.

Solving Classification Problems with daisugi

showcasing how to use daisugi for classification. our dataset comes from forested, a tabular data repo which lists forest attributes & whether an area is “forested” or “non-forested”.

Prepare Dataset

library(daisugi)
library(forested)
library(rsample)
set.seed(5311)

# defining splits for training and testing datasets
splits <- rsample::initial_split(forested::forested)
training <- rsample::training(splits)
testing <- rsample::testing(splits)


# (x, y) Training:
x_train <- training |>
  # target (and factors as not all engines handle cats)
  dplyr::select(-forested, -tree_no_tree, -land_type, -county)

# our target variable:
y_train <- training |> dplyr::select(forested) |> dplyr::pull()

# (x, y) Testing:
x_test <- testing |>
  dplyr::select(-forested, -tree_no_tree, -land_type, -county)

y_test <- testing |> dplyr::select(forested) |> dplyr::pull()

head(y_test)
#> [1] No  Yes Yes Yes Yes Yes
#> Levels: Yes No

yggdrasil decision forests

Yggdrasil Decision Forests (YDF) is Google’s high-performance tree ecosystem supporting gradient boosted trees, random forests, and specialized split strategies.

The implementation exposed through daisugi emphasizes:

• oblique random splits
• scalable forest construction
• modern decision forest infrastructure

ydf_trees <- grow_yggdrasil_trees(
  x_train,
  y_train,
  trees = 5L
)
#> Downloading uv...Done!
#> Train model on 5330 examples
#> Model trained in 0:00:00.040564

harvest_yggdrasil_trees(ydf_trees, x_test) |> head()
#> [1] "Yes" "Yes" "Yes" "Yes" "Yes" "Yes"

perpetual

Perpetual is a budget-driven boosting methodology designed around adaptive predictive scaling rather than extensive hyperparameter tuning.

The core idea is:

• increase predictive budget
• monitor loss stabilization
• stop when improvement plateaus

This creates an AutoML-like boosting workflow with minimal tuning overhead.

perpetual_trees <- grow_perpetual_trees(
  x_train,
  y_train
)

harvest_perpetual_trees(perpetual_trees, x_test) |> head()
#> [1] 0 0 0 0 0 0

wildwood

WildWood is an advanced probabilistic random forest algorithm emphasizing aggregation over multiple possible tree prunings.

Unlike standard random forests, WildWood combines:

• randomized forests
• exponential weighting
• out-of-bag pruning aggregation

This produces highly adaptive ensemble behavior.

wild_trees <- grow_wild_trees(
  x_train,
  y_train,
  trees = 5L
)

harvest_wild_trees(wild_trees, x_test) |> head()
#> [1] "Yes" "Yes" "Yes" "Yes" "Yes" "Yes"

explainable boosting machines

Explainable Boosting Machines (EBMs) are interpretable generalized additive boosting systems developed by Microsoft’s InterpretML project.

EBMs aim to balance:

• predictive performance
• transparency
• interaction discovery
• human interpretability

They are often considered “glassbox” models because their learned structure remains directly inspectable.

explainable_trees <- grow_explainable_trees(
  x_train,
  y_train,
  trees = 5L
)

harvest_explainable_trees(wild_trees, x_test) |> head()
#> [1] "Yes" "Yes" "Yes" "Yes" "Yes" "Yes"

evolutionary trees

Evolutionary Trees comes from {evtree} R package. Which involves evolutionary learning of global optimal trees for both classification and regression.

evolutionary_trees <- grow_evolutionary_trees(
  x_train,
  y_train,
  trees = 10L
)

harvest_evolutionary_trees(evolutionary_trees, x_test) |> head()
#>   1   2   3   4   5   6 
#> Yes Yes Yes Yes Yes Yes 
#> Levels: Yes No