php-ml/docs/machine-learning/classification/naive-bayes.md

NaiveBayes Classifier

Classifier based on applying Bayes' theorem with strong (naive) independence assumptions between the features.

Train

To train a classifier, simply provide train samples and labels (as array). Example:

$samples = [[5, 1, 1], [1, 5, 1], [1, 1, 5]];
$labels = ['a', 'b', 'c'];

$classifier = new NaiveBayes();
$classifier->train($samples, $labels);

You can train the classifier using multiple data sets, predictions will be based on all the training data.

Predict

To predict sample label use the predict method. You can provide one sample or array of samples:

$classifier->predict([3, 1, 1]);
// return 'a'

$classifier->predict([[3, 1, 1], [1, 4, 1]]);
// return ['a', 'b']