A=INDEP(C,H) returns a indep object initialized with density
estimator(s) C and hyperparameters H.
Training will try to fit the estimator C to each feature independently,
C can be an array of algorithms, one for each feature.
Testing will return the density estimate for each data point tested,
or if passed the empty dataset will generate new
class data according to the densities in the model.
Hyperparameters:
l=50 -- number of data points to generate if asked to generate
Model:
child={gauss} -- array of underlying density estimators
Methods:
train, test, generate
Examples:
gen(indep({gauss([-1]) gauss([1])}))
get_mean(train(cv(bayes(indep)),toy)) run naive bayes