ML.NET
Machine Learning at Microsoft with ML.NET paper
https://arxiv.org/pdf/1905.05715.pdf
https://github.com/dotnet/machinelearning
High performance and accuracy
Training on ~900 MB of an Amazon review dataset, ML.NET produced a model with 93% accuracy, scikit-learn with 92%, and H2O with 85%. ML.NET took 11 minutes to train and test the model, scikit-learn took 66 minutes, and H2O took 105 minutes.
NimbusML
https://github.com/Microsoft/NimbusML
nimbusml is a Python module that provides Python bindings for ML.NET.
Infer.NET
https://github.com/dotnet/infer
是一个在概率图模型(graphical models)中运行贝叶斯推理(Bayesian inference)的框架
AForge.NET
https://github.com/andrewkirillov/AForge.NET
AForge.NET是一个专门为开发者和研究者基于C#框架设计的,他包括计算机视觉与人工智能,图像处理,神经网络,遗传算法,机器学习,模糊系统,机器人控制等领域。这个框架由一系列的类库组成。主要包括有:
AForge.Imaging —— 一些日常的图像处理和过滤器
AForge.Vision —— 计算机视觉应用类库
AForge.Neuro —— 神经网络计算库AForge.Genetic -进化算法编程库
AForge.MachineLearning —— 机器学习类库
AForge.Robotics —— 提供一些机器学习的工具类库
AForge.Video —— 一系列的视频处理类库
AForge.Fuzzy —— 模糊推理系统类库
AForge.Controls—— 图像,三维,图表显示控件
Accord.NET
https://github.com/accord-net/framework
Accord.NET Framework是在AForge.NET基础上封装和进一步开发来的。功能也很强大,因为AForge.NET更注重与一些底层和广度,而Accord.NET Framework更注重与机器学习这个专业,在其基础上提供了更多统计分析和处理函数,包括图像处理和计算机视觉算法,所以侧重点不同,但都非常有用。
绑定
https://github.com/SciSharp/TensorFlow.NET
https://github.com/migueldeicaza/TensorFlowSharp
https://github.com/SciSharp/Torch.NET
https://github.com/xamarin/TorchSharp
https://github.com/tech-quantum/MxNet.Sharp
https://github.com/SciSharp/Keras.NET
https://github.com/SciSharp/SharpCV
https://github.com/shimat/opencvsharp
SciSharp其它生态
https://github.com/SciSharp/NumSharp
https://github.com/SciSharp/Plot.NET
https://github.com/SciSharp/Pandas.NET
https://github.com/SciSharp/PillowSharp
https://github.com/SciSharp/Matplotlib.Net
https://github.com/SciSharp/Gym.NET
其它机器学习库
https://github.com/sethjuarez/numl
https://github.com/mdabros/SharpLearning
https://github.com/jdermody/brightwire
Spark绑定
https://github.com/dotnet/spark
自动驾驶仿真器
https://github.com/lgsvl/simulator
人脸识别
https://github.com/takuya-takeuchi/FaceRecognitionDotNet
OpenPose人体姿态识别
https://github.com/takuya-takeuchi/OpenPoseDotNet
神经网络
https://github.com/Sergio0694/NeuralNetwork.NET
DL
https://github.com/harujoh/KelpNet
https://github.com/kawatan/Merkurius
RNN
https://github.com/zhongkaifu/RNNSharp
NLP
https://github.com/curiosity-ai/catalyst
https://github.com/SciSharp/CherubNLP
开源.NET平台非综合类
Math.NET
Math.NET是.NET平台下最全面的数学计算组件之一,基础功能非常完善。
Adaboost算法
1.https://github.com/bgorven/Classifier
2.https://github.com/ElmerNing/Adaboost
Apriori算法
1.https://github.com/Omar-Salem/Apriori-Algorithm
2.https://github.com/simonesalvo/apriori
PageRank算法
https://github.com/archgold/pagerank
NativeBayes(朴素贝叶斯)算法
1.https://github.com/Rekin/Naive-Bayes-Classifier
2.https://github.com/ArdaXi/Bayes.NET
3.https://github.com/amrishdeep/Dragon
4.https://github.com/joelmartinez/nBayes
kmeans算法
http://visualstudiomagazine.com/articles/2013/12/01/k-means-data-clustering-using-c.aspx