I'm looking for an implementation of Similarity Forest algorithm as defined by Sathe et al. in their KDD paper (https://www.kdd.org/kdd2017/papers/view/similarity-forests). So far, I have found this one in Python https://github.com/rrricharrrd/similarity-forest, but the author states himselves that his version is "Basic (not especially-optimised)". I do not get good results with this version of the algorithm for my specific study and I wonder if I would get better results with more opportunities of parameter tuning. Would other versions be available in R or Python ?


I implemented this algorithm in Python: https://github.com/sfczekalski/similarity_forest

My implementation includes not only binary classification, as in the paper, but also multiclass classification, regression, anomaly detection and metric learning for clustering. Some example codes are provided in examples folder.

My work is still in progress, more documentation, examples, and code improvements will be added.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.