PMSE: Predictions and Mean squared error

In kusakehan/TEReg: An R package for tensor envelope regression with tensor response and tensor predictor

Description

Evaluate tensor predictor regression through prediction mean squared error.

Usage

PMSE(Xn, Yn, Bhat)

Arguments

Xn	A predictor tensor.
Yn	A response vector.
Bhat	An estimation of coefficient tensor.

Value

mse	Mean squared error. Defined as trace(∑_{i=1}^n(\mathbf{Y}_i-\hat{\mathbf{Y}}_i)(\mathbf{Y}_i-\hat{\mathbf{Y}}_i)'/n), where \hat{\mathbf{Y}}_i is the predictions.
Yhat	The predictions of tensor predictor regression.

kusakehan/TEReg documentation built on May 30, 2019, 7:17 a.m.