Partial port of scikit-learn to go

DBSCAN KMeans

LoadIris LoadBreastCancer LoadDiabetes LoadBoston LoadExamScore LoadMicroChipTest LoadMnist LoadMnistWeights MakeRegression MakeBlobs

CubicSpline Interp1d Interp2d

ConstantKernel WhiteKernel RBF DotProduct

LinearRegression BayesianRidge MultiTaskElasticNet MultiTaskLasso ElasticNet Lasso LassoPath LogisticRegression Ridge

AccuracyScore ConfusionMatrix PrecisionScore RecallScore F1Score FBetaScore PrecisionRecallFScoreSupport ROCCurve AUC ROCAUCScore PrecisionRecallCurve AveragePrecisionScore R2Score

KFold CrossValidate

KNeighborsClassifier MinkowskiDistance EuclideanDistance KDTree NearestCentroid KNeighborsRegressor NearestNeighbors NearestNeighbors.KNeighborsGraph NearestNeighbors.Tree

MLPClassifier.Unmarshal MLPClassifier.Fit.mnist MLPClassifier.Predict.mnist MLPClassifier.Fit.breast.cancer MLPRegressor.Fit.boston

Pipeline

MinMaxScaler StandardScaler RobustScaler AddDummyFeature OneHotEncoder Shuffler MaxAbsScaler Binarizer Normalizer Scale KernelCenterer QuantileTransformer PowerTransformer PowerTransformer.boxcox KBinsDiscretizer FunctionTransformer Imputer LabelBinarizer MultiLabelBinarizer LabelEncoder PCA

SVC SVR

This is a personal project to get a deeper understanding of how all of this magic works

• linted with gofmt, golint, go vet revive

• unit tested but coverage should reach 90%

• underdocumented but scikit-learn doc is your friend

Many thanks to gonum and scikit-learn authors and contributors

PRs are welcome