Machine Learning with Swift

AIToolbox

The Swift library contains multiple machine learning models. They are compatible with both iOS and macOS. All models are implemented as separate classes with unified interfaces, so you can replace one model with another in your code with minimal effort. Some models support saving to plist files and loading from such files. The Accelerate framework is used throughout the library to boost the speed of calculations.

For regression tasks, you can choose between linear, nonlinear, and SVM regression. Linear regression supports regularization. The SVM model here is a port of the libSVM library initially written in C and can also be used for classification. Other classification algorithms include logistic regression and neural networks. ...