Classification pipeline functions
A monadic pipeline is a way to navigate computations in monad's function space. The unit function and the setter functions "uplift" data into monad's space. The taker functions "download" data from monad's space. The central -- and most computationally intensive -- functions are the classifier training and measurement functions. Variety of use cases are addressed with the rest of the functions. Diagnosis with Receiver Operating Characteristic (ROC) functions are extensively supported.
Fundamental
— monad objects head
— create monad unit object
— monad's failure symbol
— check whether an expression is a monad object
— monad's binding function
Setters
Takers
Removers
Preliminary
Classification
— make a classifier (with given algorithm names and options)
— suggest classification thresholds based on ROC data
Classifier measurements
— computes specified measurements
— computes specified measurements using a threshold for a specified class label
— computes ROC plot data
— plots ROC curves
Importance of variables
— computes variable importance
""