Implement UMAP - Githubissues

h2oai / h2o-3

H2O is an Open Source, Distributed, Fast & Scalable Machine Learning Platform: Deep Learning, Gradient Boosting (GBM) & XGBoost, Random Forest, Generalized Linear Modeling (GLM with Elastic Net), K-Means, PCA, Generalized Additive Models (GAM), RuleFit, Support Vector Machine (SVM), Stacked Ensembles, Automatic Machine Learning (AutoML), etc.

http://h2o.ai

Apache License 2.0

6.78k stars 1.99k forks source link

Implement UMAP #16271

Open wendycwong opened 1 month ago

wendycwong commented 1 month ago

From paid customer.

Here is one link: https://umap-learn.readthedocs.io/en/latest/

To complete this task, apart from implementing the algorithm and mojo, please also complete a paper describing the actual implementation. This paper can be used by co-workers in the code review process and help the documentation team as well.

I am attaching an example here for when I implemented GAM: GAM_doc.pdf

If you can divide your task into meaningful and independent parts, feel free to farm them out to others if you need the help. I for one may be able to help with coming up with the paper/doc.

maurever commented 4 days ago

The paper describing the H2O.ai implementation.

Implementation plan

Understand UMAP implementation (in progress)
Understand the theory behind (in progress)
Think about parallelization - decide if it is possible (in progress)
Think about time estimate - decide the time we need (in progress)
Decide if implement KNN as a model or as a part of UMAP (in progress)
Implement KNN
Implement matrices operations
Implement the first version of UMAP
- Java backend with minimal parameters as possible
- JUnit tests
- Documentation
- MOJO
- Python/R API
- Python/R tests
- Performance analysis
- Compare results with original UMAP implementation