naive-bayes
naive-bayes copied to clipboard
mljs
Naive bayes classifier.
If you have a large dataset, of perhaps 5 GB+, then it is not efficient to do an in-memory `model.train(sets, labels)` approach whereas `sets` and `labels` could consist of hundreds...
When attempting to run with the following data-sets: ``` const cases = [ [ 200, 100, 1500, 500, 2700 ], [ 300, 25, 1555, 750, 2800 ], [ 150, 200,...
My fix for this issue that was least invasive. Check the additional tests in `/GaussianNB`, it now supports a lot of chars, numbers, etc. as prediction symbols, not only consecutive...
When processing arbitrary statistical data, it is possible that some columns just may have a very low variance for one or more classes being trained. While it might be desirable...
When placing my sample data (see below) into GaussianNB 'Native bayes' / 'Basic test', the test fails and all results only belong to a single class. What may be the...
As stated in the title, I would like to suggest to add a new optional parameter in GaussianNB called sample_weight, similar to sklearn. `nb.train(x, y[, sample_weight]) `
