irithyll 10.0.0

Streaming ML in Rust -- gradient boosted trees, neural architectures (TTT/KAN/MoE/Mamba/SNN), AutoML, kernel methods, and composable pipelines
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

//! Streaming learner implementations for polymorphic model composition.
//!
//! Each learner in this module implements the [`crate::learner::StreamingLearner`] trait,
//! enabling runtime-polymorphic stacking via `Box<dyn StreamingLearner>`.
//!
//! # Learners
//!
//! | Module | Type | Algorithm |
//! |--------|------|-----------|
//! | [`linear`] | [`StreamingLinearModel`] | SGD with L1/L2/ElasticNet regularization |
//! | [`naive_bayes`] | [`GaussianNB`] | Incremental Gaussian Naive Bayes classifier |
//! | [`mondrian`] | [`MondrianForest`] | Online random forest with lifetime-based splits |
//! | [`rls`] | [`RecursiveLeastSquares`] | Exact streaming OLS via Sherman-Morrison |
//! | [`rls`] | [`StreamingPolynomialRegression`] | RLS with polynomial feature expansion |
//! | [`rls`] | [`LocallyWeightedRegression`] | Nadaraya-Watson with circular buffer |
//! | [`krls`] | [`KRLS`] | Kernel RLS with ALD sparsification |

pub mod bounded_rls;
pub mod classification;
pub mod krls;
pub mod linear;
pub mod mondrian;
pub mod multinomial_bernoulli_nb;
pub mod naive_bayes;
pub mod rls;

pub use bounded_rls::BoundedRls;
pub use classification::{ClassificationMode, ClassificationWrapper};
pub use krls::{Kernel, LinearKernel, PolynomialKernel, RBFKernel, KRLS};
pub use linear::StreamingLinearModel;
pub use mondrian::MondrianForest;
pub use multinomial_bernoulli_nb::{BernoulliNB, MultinomialNB};
pub use naive_bayes::GaussianNB;
pub use rls::{LocallyWeightedRegression, RecursiveLeastSquares, StreamingPolynomialRegression};