ml_algo library

Classes

CrossValidator: A factory and an interface for all the cross validator types
DecisionTreeClassifier: A class that performs decision tree-based classification
KDTree: KD-tree - an algorithm that provides efficient data retrieval by splitting the whole searching space into partitions in form of binary tree which means that data querying on average will take O(log(n)) time
KnnClassifier: A class that performs classification based on k nearest neighbours (KNN) algorithm
KnnRegressor: A class that performs regression basing on k nearest neighbours algorithm
LinearRegressor: Linear regression
LogisticRegressor: Logistic regression-based classification.
RandomBinaryProjectionSearcher: Random Binary Projection is a an algorithm that randomly partitions all reference data points into different bins, which makes it possible to perform efficient K Nearest Neighbours search, since there is no need to search for the neighbours through the entire data: it's needed to visit just a few bins to look for the neighbours.
SoftmaxRegressor: Softmax classification

Distance: A type of vector distance
DType: A type, denoting, how data will be stored and processed in vectors or matrices
InitialCoefficientsType: Defines the nature of coefficients for linear models at the beginning of fitting
KDTreeSplitStrategy
KernelType: A type of a kernel function.
LearningRateType: A type of learning rate strategies
LinearOptimizerType: Linear optimization types
MetricType: Metrics for measuring the quality of the prediction.
RegularizationType: Regularization types
TreeAssessorType

splitData(DataFrame data, Iterable<double> ratios) → List<DataFrame>: Splits the given data into parts depending on ratios. The number of the produced parts can be described as ratios.length + 1

DataPreprocessFn = List<DataFrame> Function(DataFrame trainData, DataFrame testData)
ModelFactory = Assessable Function(DataFrame observations)