From 6c93efe04f266301bffa308847a785f91b48cde1 Mon Sep 17 00:00:00 2001 From: Gal Oshri Date: Sun, 6 May 2018 11:42:09 -0700 Subject: [PATCH 1/2] Add release notes for ML.NET 0.1 --- .../release-notes/0.1/release-0.1.md | 39 +++++++++++++++++++ 1 file changed, 39 insertions(+) create mode 100644 Documentation/release-notes/0.1/release-0.1.md diff --git a/Documentation/release-notes/0.1/release-0.1.md b/Documentation/release-notes/0.1/release-0.1.md new file mode 100644 index 0000000000..f0900424b5 --- /dev/null +++ b/Documentation/release-notes/0.1/release-0.1.md @@ -0,0 +1,39 @@ +# ML.NET 0.1 Release Notes + +ML.NET 0.1 is the first preview release of ML.NET. Thank you for trying it out and we look forward to your feedback! + +### Installation + +You can install ML.NET NuGet from the CLI using: +``` +dotnet add package Microsoft.ML +``` + +From package manager: +``` +Install-Package Microsoft.ML +``` + +### Release Notes + +This initial release contains core ML.NET components for enabling machine learning pipelines: + +* ML Data Structures (e.g. `IDataView`, `LearningPipeline`) +* TextLoader (loading data from a text file into a `LearningPipeline`) +* Transforms (to get data in the correct format for training): + * Processing/featurizing text: `TextFeaturizer` + * Schema modifcation: `ColumnConcatenator`, `ColumnSelector`, and `ColumnDropper` + * Working with categorical features: `CategoricalOneHotVectorizer` and `CategoricalHashOneHotVectorizer` + * Dealing with missing data: `MissingValueHandler` + * Filters: `RowTakeFilter`, `RowSkipFilter`, `RowRangeFilter` + * Feature selection: `FeatureSelectorByCount` and `FeatureSelectorByMutualInformation` +* Learners (to train machine learning models) for a variety of tasks: + * Binary classification: `FastTreeBinaryClassifier`, `StochasticDualCoordinateAscentBinaryClassifier`, `AveragedPerceptronBinaryClassifier`, `BinaryLogisticRegressor`, `FastForestBinaryClassifier`, `LinearSvmBinaryClassifier`, and `GeneralizedAdditiveModelBinaryClassifier` + * Multiclass classification: `StochasticDualCoordinateAscentClassifier`, `LogisticRegressor`, and`NaiveBayesClassifier` + * Regression: `FastTreeRegressor`, `FastTreeTweedieRegressor`, `StochasticDualCoordinateAscentRegressor`, `OrdinaryLeastSquaresRegressor`, `OnlineGradientDescentRegressor`, `PoissonRegressor`, and `GeneralizedAdditiveModelRegressor` +* Evaluators (to check the model works well): + * For Binary classification: `BinaryClassificationEvaluator` + * For Multiclass classification: `ClassificationEvaluator` + * For Regression: `RegressionEvaluator` + +Additional components have been included in the repository but cannot be used in `LearningPipeline` yet (this will be updated in future releases). \ No newline at end of file From 3530b523b3ee944a0d5b004e4ab015dea55fb8ae Mon Sep 17 00:00:00 2001 From: danmosemsft Date: Sun, 6 May 2018 21:03:43 -0700 Subject: [PATCH 2/2] Release note tweaks --- Documentation/release-notes/0.1/release-0.1.md | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-) diff --git a/Documentation/release-notes/0.1/release-0.1.md b/Documentation/release-notes/0.1/release-0.1.md index f0900424b5..a36055527a 100644 --- a/Documentation/release-notes/0.1/release-0.1.md +++ b/Documentation/release-notes/0.1/release-0.1.md @@ -1,10 +1,12 @@ # ML.NET 0.1 Release Notes -ML.NET 0.1 is the first preview release of ML.NET. Thank you for trying it out and we look forward to your feedback! +ML.NET 0.1 is the first preview release of ML.NET. Thank you for trying it out and we look forward to your feedback! Try training, scoring, and using machine learning models in your app and tell us how it goes. ### Installation -You can install ML.NET NuGet from the CLI using: +ML.NET works on any platform that supports [.NET Core 2.0](https://www.microsoft.com/net/learn/get-started/windows). It also works on the .NET Framework. + +You can install ML.NET NuGet from the .NET Core CLI using: ``` dotnet add package Microsoft.ML ``` @@ -14,12 +16,16 @@ From package manager: Install-Package Microsoft.ML ``` +Or from within Visual Studio's NuGet package manager. + ### Release Notes This initial release contains core ML.NET components for enabling machine learning pipelines: * ML Data Structures (e.g. `IDataView`, `LearningPipeline`) -* TextLoader (loading data from a text file into a `LearningPipeline`) + +* TextLoader (loading data from a delimited text file into a `LearningPipeline`) + * Transforms (to get data in the correct format for training): * Processing/featurizing text: `TextFeaturizer` * Schema modifcation: `ColumnConcatenator`, `ColumnSelector`, and `ColumnDropper` @@ -27,13 +33,15 @@ This initial release contains core ML.NET components for enabling machine learni * Dealing with missing data: `MissingValueHandler` * Filters: `RowTakeFilter`, `RowSkipFilter`, `RowRangeFilter` * Feature selection: `FeatureSelectorByCount` and `FeatureSelectorByMutualInformation` + * Learners (to train machine learning models) for a variety of tasks: * Binary classification: `FastTreeBinaryClassifier`, `StochasticDualCoordinateAscentBinaryClassifier`, `AveragedPerceptronBinaryClassifier`, `BinaryLogisticRegressor`, `FastForestBinaryClassifier`, `LinearSvmBinaryClassifier`, and `GeneralizedAdditiveModelBinaryClassifier` * Multiclass classification: `StochasticDualCoordinateAscentClassifier`, `LogisticRegressor`, and`NaiveBayesClassifier` * Regression: `FastTreeRegressor`, `FastTreeTweedieRegressor`, `StochasticDualCoordinateAscentRegressor`, `OrdinaryLeastSquaresRegressor`, `OnlineGradientDescentRegressor`, `PoissonRegressor`, and `GeneralizedAdditiveModelRegressor` -* Evaluators (to check the model works well): + +* Evaluators (to check how well the model works): * For Binary classification: `BinaryClassificationEvaluator` * For Multiclass classification: `ClassificationEvaluator` * For Regression: `RegressionEvaluator` -Additional components have been included in the repository but cannot be used in `LearningPipeline` yet (this will be updated in future releases). \ No newline at end of file +Additional components have been included in the repository but cannot be used in the `LearningPipeline` yet (this will be updated in future releases). \ No newline at end of file