aom_ridge_blender — n4m.ensemble.aom_ridge_blender

Namespace: n4m.ensemble · Fully-qualified: n4m.ensemble.aom_ridge_blender · Catalog id: aom_pop.ridge_blender

API surface

C ABI (ABI 2): n4m_ensemble_aom_ridge_blender_fit. Use the linked public header for the exact signature, configuration, and result handles.

Python (verified public re-export): from n4m.ensemble import aom_ridge_blender

Signature: aom_ridge_blender(X, y, *, profile: str | int = 'compact', cv: int = 5, fold_ids = None, ridge_lambdas = (0.0001, 0.01, 1.0, 100.0), regularizer: float = 0.01, center_x: bool | None = None, scale_x: bool | None = None, center_y: bool | None = None, scale_y: bool | None = None)

R: no current source-verified entry point was found for this catalog method.

MATLAB / Octave: no current source-verified entry point was found for this catalog method.

Parameters

Name

Type

Default

X

—

required

y

—

required

profile

str | int

'compact'

cv

int

5

fold_ids

—

None

ridge_lambdas

—

(0.0001, 0.01, 1.0, 100.0)

regularizer

float

0.01

center_x

bool | None

None

scale_x

bool | None

None

center_y

bool | None

None

scale_y

bool | None

None

Explanations

Bibliographic source

No canonical paper defines this exact n4m candidate bank; it uses non-negative stacking. Beurier, G. et al. (2026). AOM-PLS / POP-PLS paper companion, arXiv:2605.13587, https://arxiv.org/abs/2605.13587. The product variants below are implementation-specific extensions; the paper does not by itself specify their ABI-2 orchestration.

Mathematical principle

Generate OOF predictions for every strict-linear AOM Ridge candidate, solve non-negative weights constrained to sum to one, refit each candidate on all rows, and blend their raw-space linear coefficients.

Appropriate uses

Stable combination of several plausible strict AOM Ridge candidates while retaining a replayable linear predictor.

Limits and validation

Weights must be learned from OOF rather than in-sample predictions. Highly correlated candidates can yield unstable weights; the native bank is finite and is not a generic arbitrary-estimator blender.

Implementation

n4m.ensemble.aom_ridge_blender and AOMRidgeBlenderRegressor; C ABI n4m_ensemble_aom_ridge_blender_fit.

Sources and provenance

https://github.com/GBeurier/nirs4all-methods/blob/main/cpp/src/core/aom_ridge_blender.cpp; https://github.com/GBeurier/nirs4all-methods/blob/main/bindings/python/src/n4m/ensemble.py

Catalog note

Native strict-linear AOM Ridge OOF simplex blender over compact/wide chain banks and positive Ridge-lambda grids. Compact has 12 chains; wide has 31 strict-linear chains, including Gaussian, FCK and Whittaker variants. It returns candidate predictions, OOF candidate predictions, non-negative blend weights, and final input_coefficients plus intercept for replay through NativeAOMRidgeBlenderRegressor; native v1 builds in CUDA-enabled configurations but is not yet a fused batched GPU blender.

Timing benchmark: benchmarks/cross_binding/bench_aom_ridge_blender_timing.py

See also: methods index.