Train Regressor (GLM / Tweedie) Node

T

AI/ML/Regression

Fit/Train a Generalized Linear Model. Pick the distribution that matches the target: Normal for unbounded values, Poisson for counts, Gamma for positive skewed amounts, Inverse Gaussian for heavy tails.

fit_glmml

Inputs8

Outputs2

Security exposure6/10

Packageml

Ratings

Scores range from 0 to 10. Higher values mean more impact, exposure, or operational weight.

SecurityAttack surface and exposure impact.

6/10Medium

PrivacyPotential sensitivity of processed data.

6/10Medium

PerformanceRuntime or resource pressure.

6/10Medium

GovernancePolicy, audit, or compliance impact.

8/10High

ReliabilityOperational stability considerations.

7/10High

CostExternal or compute cost impact.

7/10High

Input Pins

8

Input

Execution

exec_in

Execution trigger that begins GLM training

Data Source

String

source

Choose which backend supplies the training data

Default Database

Database

Distribution

String

distribution

Target distribution: Normal (power 0, any value), Poisson (power 1, counts >= 0), Gamma (power 2, values > 0), Inverse Gaussian (power 3, values > 0), or Custom to set the Tweedie power directly

Default Normal

NormalPoissonGammaInverse GaussianCustom

Tweedie Power

Float

power

Free Tweedie power, only used when Distribution is Custom. Values in (0, 1) do not describe any distribution and are rejected; (1, 2) is compound Poisson-Gamma.

Default 0

Range -5 to 5

Alpha (L2 Penalty)

Float

alpha

Strength of the L2 penalty on the coefficients. 0 fits an unpenalized GLM.

Default 1

Range 0 to 1000

Fit Intercept

Boolean

fit_intercept

Fit a bias term. Disable only when the data is already centered.

Default true

Max Iterations

Integer

max_iter

Iteration cap for the L-BFGS solver. Defaults to 1000 instead of the library default of 100, which is too low to converge on unscaled real-world features.

Default 1000

Range 1 to 1000000

Tolerance

Float

tol

Gradient tolerance that stops the L-BFGS solver. Smaller values fit tighter but need more iterations.

Default 0.0001

Range 1e-12 to 1

Output Pins

2

Done

Execution

exec_out

Activated once training completes

Model

Struct

model

Thread-safe handle to the trained generalized linear model

SchemaNodeMLModelNodeMLModelNodeMLModel1 fields

model_refstringrequired

Schema enforced

Node Info

Internal name

fit_glm

Category

AI/ML/Regression