Module decision_tree_learner 

DecisionTreeLearner — ID3/C4.5-style decision tree with training, prediction, feature importance, pruning, and rich statistics.

Design

• Recursive binary splitting on continuous features.
• Configurable split criterion: Shannon Entropy (ID3), Gini impurity, or Misclassification rate.
• Optional feature sub-sampling per split node (random forests style), driven by a pure-Rust xorshift64 PRNG — no external crate required.
• Post-training leaf pruning (prune) to collapse subtrees where both descendant leaves are minority-dominated.
• Bounded training-history ring-buffer (VecDeque capped at 100 records).
• All operations are no_std-compatible except for HashMap/VecDeque.

Structs

DecisionTreeLearner

ID3/C4.5-style decision tree learner.

DtlLearnerConfig

Hyper-parameters controlling tree growth and prediction.

DtlLearnerStats

Summary statistics produced by DecisionTreeLearner::learner_stats.

DtlPrediction

Prediction result from DecisionTreeLearner::predict.

DtlSample

One labeled training / prediction example.

DtlTrainingRecord

One entry in the training history ring-buffer.

Enums

DtlCriterion

Split criterion for tree growing.

DtlError

Errors produced by DecisionTreeLearner.

DtlNode

Internal node of the decision tree.

Type Aliases

DtlDecisionTreeLearner

Alias for DecisionTreeLearner.