MATLAB File Help: dml.glmnet

  glmnet efficient elastic net algorithm.
 
    DESCRIPTION
    Fast implementation of elastic net linear and logistic regression by
    Friedman et al. Without arguments, this method automatically determines 
    the lambda path and uses cross-validation to find the best lambda. The
    parameter 'family' determines whether to use linear regression
    (gaussian) or logistic regression (binomial or multinomial). The mixing
    parameter alpha controls the influence of the 11 (alpha=1) and l2 regularizer
    (alpha = 0).
 
    REFERENCE
    Regularization paths for generalized linear models via coordinate descent 
    by Friedman et al.
 
    EXAMPLE
    X = rand(10,20); Y = [1 1 1 1 1 2 2 2 2 2]';
    m = dml.glmnet('family','binomial');
    m = m.train(X,Y);
    Z = m.test(X);
 
    X = rand(15,20); Y = [1 1 1 1 1 2 2 2 2 2 3 3 3 3 3]'; X(1:5,:) = X(1:5,:)+1;
    m = dml.glmnet('family','multinomial');
    m = m.train(X,Y);
    Z = m.test(X);
 
    DEVELOPER
    Marcel van Gerven (m.vangerven@donders.ru.nl)

Superclasses	dml.method
Sealed	false
Construct on load	false

alpha	glmnet mixing parameter
df	degrees of freedom
family	gaussian, binomial, or multinomial
indims	dimensions of the input data (excluding the trial dim and time dim in time series data)
lambda	used lambda(s)
path	lambda path followed when estimating decoding performance
performance	cross-validated decoding performance for elements on lambda path
restart	when false, starts at the previously learned parameters; needed for online learning and grid search
validator	cross-validator
verbose	whether or not to generate diagnostic output
weights	regression weights (offset last)

	model	returns
	test
	train	handle multiple datasets