svr_poly – Support vector regression with polynomial kernel.#

Back to family axis | Back to L4 | Browse all options

Operational op under axis family, sub-layer L4_A_model_selection, layer l4. Standalone callable: mf.functions.svr_poly_fit.

Function signature#

mf.functions.svr_poly_fit(
    X: np.ndarray | pd.DataFrame,
    y: np.ndarray | pd.Series,
) -> SVRPolyFitResult

Parameters#

name

type

default

constraint

description

X

`np.ndarray

pd.DataFrame`

—

—

y

`np.ndarray

pd.Series`

—

—

Returns#

SVRPolyFitResult — frozen dataclass with fit results.

Attribute

Type

Description

.C

float

Regularisation parameter used.

.degree

int

Polynomial degree.

.n_support_vectors

int

Number of support vectors.

.predict(X)

np.ndarray

Predictions for new data X, shape (n_samples,).

.summary()

str

Table: C, degree, support vector count.

Behavior#

Polynomial-kernel SVR. Useful for studies that want explicit polynomial features without manual expansion.

When to use

Polynomial-kernel ablations. Selecting svr_poly on l4.family activates this branch of the layer’s runtime.

In recipe context#

Set params.family = "svr_poly" in the relevant layer to activate this op within a recipe:

# Layer L4 recipe fragment
params:
  family: svr_poly

References#

  • macroforecast design Part 2, L4: ‘forecasting model is the layer where every authoring iteration ends – pick family, tune, repeat.’