Andrew DalPino

Hi @markrogoyski @Beakerboy Are there plans for implementing QR still? It would be SUPER helpful to me and I'm sure others in the case where I need eigenvalues and eigenvectors...

Hey @GeorgeGardiner great question ... summarizing our convo from the Telegram Channel (https://t.me/RubixML) ... You can use the transformer pipeline from the [Sentiment example](https://github.com/RubixML/Sentiment) with [Gaussian Naive Bayes](https://docs.rubixml.com/en/latest/classifiers/gaussian-naive-bayes.html) under the...

Hi @nizariyah thanks for the question [Classification Tree](https://docs.rubixml.com/en/latest/classifiers/classification-tree.html) minimizes Gini impurity and [Extra Tree Classifier](https://docs.rubixml.com/en/latest/classifiers/extra-tree-classifier.html) minimizes entropy at the leaf nodes of the tree. For regression, both [Regression Tree](https://docs.rubixml.com/en/latest/regressors/regression-tree.html) and...

Hi @nizariyah The CART implementation has been tuned and optimized in the latest commit https://github.com/RubixML/RubixML/commit/89f6991794c9ee5e7a2f358c1cc52167450684a8 Instead of using a 0.25 ratio of quantiles to samples at the split node, we...

Performance comparison of PHP library Random Forest on the Dota 2 dataset, 10 tree, 0.5 subsample ratio, categorical features  

Great question @marclaporte! I am not expert but I read the document you linked and it looks like they're performing some type of full-text search using TF-IDF (Term Frequency Inverse...

Hey @alaindesilets Yes you'd increase *k* with the number of similar documents you wanted retrieved and instead of the `predict()` API you'd use `proba()` to return an array of probabilities....

We added maximum document frequency to [Word Count Vectorizer](https://docs.rubixml.com/en/latest/transformers/word-count-vectorizer.html) in the last commit (https://github.com/RubixML/RubixML/commit/52033972b1193671b6e4c1a4d005918c5a825e27) so now you can narrow the vocabulary by removing frequent tokens similarly to Elasticsearch. **Example** ```php...

Great @marclaporte I already began doing some experiments on my end Victor and I will take it from there when ready

Here is what I came up with - it uses the positive reviews from the [Sentiment](https://github.com/RubixML/Sentiment) dataset. You should be able to drop this code right into the project (ex...