Releases: rafaelgreca/scratchml
Releases · rafaelgreca/scratchml
Releasing v10.1.1
Updating requirements and reducing tests repetitions.
v10.1.0
Rolling back and fixing One Hot Encoder implementation.
Releasing v10.0.0
Fixing binary labeling problem in perceptron model (Issue #19)
Adding the SVR code implemented by @MatusPohorenec (Issue #17)
Adding the SVC code implemented by @Renkooo (Issue #14)
Adding the MLP Regression code implemented by @SebastianOndrus (Issue #15)
Updating docs and README
Updating and fixing CI unit test workflow (Issue #16)
Making ScratchML available to the public!
What you will see in this release:
- Implementation of some ML algorithms (note: MLP is not ready for production yet))
- Broad test coverage;
- A lot of activation functions and accuracy metrics (both for regression and classification) to use;
- Implementation of the Binary Cross Entropy and Cross Entropy functions;
- A lot of distance metrics to be used in clustering algorithms;
- Implementation of the Label Encoder and One-hot Encoding scalers;
- Implementation of the traditional Train-Test-Split data split function, alongside KFold, Stratified KFold, and Split Into Batches.