log pi_k + sum_j log N(x_j; mu_kj, v_kj) with maximum-likelihood class
variances plus var_smoothing (ESL eq 6.26); classify to the largest.
Usage
morie_esl_naive_bayes(X, y, query = NULL, var_smoothing = 1e-09)
Arguments
- X
Training predictors.
- y
Class labels.
- query
Points to score (default X).
- var_smoothing
Added to every variance.
Value
Named list: estimate, prediction, log_posterior (query by K),
classes, priors, n, p, K.
References
Hastie, Tibshirani & Friedman (2009), sec. 6.6.3.
Examples
i <- 1:30
morie_esl_naive_bayes(cbind(sin(i), cos(3 * i) + (i %% 3)), i %% 3, rbind(c(0.2, 1)))$prediction
#> [1] 1
