Interaction settings¶
These parameters control which interactions the trees build. The highest interaction order is
set by max_interaction_order and the resolution of the
trees by max_depth. See
Choosing the tree structure for details.
interaction_gain_hurdle¶
Description¶
The interaction hurdle. While a tree grows, a split on a feature the tree does not use yet raises the tree's interaction order. Such a split is selected only if its gain is large enough relative to the gain of the tree's first split, and if it beats the best split on a feature the tree already uses. Otherwise the tree keeps refining the features it already has. This is soft heredity: interactions are admitted only on real evidence. A split that would raise the tree's order above 3 faces a doubled hurdle.
The value is the strength of the hurdle and must be finite and non-negative. 0 turns the hurdle off: each level takes the best split on a feature the tree does not use yet, and splits again on a feature it already uses only when no new feature is available.
Type
float
Default value
2.0
interaction_gain_hurdle_mode¶
Description¶
How interaction_gain_hurdle is applied.
Possible values:
adaptive— The hurdle starts at full strength and relaxes as the gains of the first splits fade during the training. Three-way interactions face a stricter hurdle than two-way ones, and must also be supported by evidence for their pairs.fixed— The hurdle is applied as given throughout the training.
Type
string
Default value
adaptive
table_budget_cells¶
Description¶
The cell budget of the table-size prior, which steers the trees toward smaller tables.
When the splits are ranked, the score of a candidate is multiplied by
where \(cells\) is the number of cells of the table the split would contribute to (the product of the numbers of cells of its features). The prior never rejects a split and never caps the size of an exported table.
Type
int
Default value
None. Depends on the tree structure:
max_depthandmax_interaction_orderboth at most 3: 2 000 000 (effectively no prior)- otherwise: 4096
table_budget_order_shrink¶
Description¶
How much table_budget_cells shrinks for each interaction order above 3.
A \(k\)-way table with \(k > 3\) is measured against
\(\frac{budget}{table\_budget\_order\_shrink^{k - 3}}\) cells, because a table with more axes is
harder to read at the same number of cells. 1 turns the shrinking off.
Type
float
Default value
2.0