fastRidgeRegression: Fast Ridge Regression with Training and Test Data

In WeakCha/BIOSTAT625_HW4: Linear Regression

Description

Fast Ridge Regression with Training and Test Data, it will return the prediction result of the data X_test and the RMSE between Y_test and the prediction results.

Usage

fastRidgeRegression(X_train, Y_train, X_test, Y_test, training_part = 0.8)

Arguments

X_train	The data used for training the linear regression model, with only features.
Y_train	The data used for training the linear regression model, with only target.
X_test	The data used for prediction and computing RMSE, with only features.
Y_test	The data used for prediction and computing RMSE, with only target.
training_part	training_part is the ratio for training data.

Value

pred	The prediction for the test set the algorithm automatically splitted
RMSE	The RMSE between the prediction and true target result for the test set

Author(s)

Li (Richard) Liu

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (X_train, Y_train, X_test, Y_test, training_part = 0.8)
{
    res_svd = svd(X_train)
    tuy = crossprod(res_svd$u, Y_train)
    beta = res_svd$v %*% (tuy * res_svd$d/(res_svd$d^2))
    pred = crossprod(t(X_test), beta)
    RMSE = eval_metrics(Y_test, pred)
    return(list(pred = pred, RMSE = RMSE))
  }

WeakCha/BIOSTAT625_HW4 documentation built on Dec. 18, 2021, 7:16 p.m.