Auto Train ML.Net Model and Get Multiple Outputs RRS feed

  • Question

  • Hi All,

    I am very new to ML.Net framework, however looking at the videos and referring to articles I could able to learn something, I explored the following question over the internet, but I couldn’t find a proper solution. If anyone can help it is great help to me.

    1. I found a code sample to Train a data-set for giving by a particular time slot

        private static void TrainData()
              MLContext mlContext = new MLContext();
               var trainingData = lContext.Data.LoadFromTextFile<SalaryData> 
               (_dataFile, hasHeader: true, separatorChar: ',');
               var settings = new RegressionExperimentSettings()
                     MaxExperimentTimeInSeconds = 20
               var experiment = 
               var result = experiment.Execute(trainingData, 
               labelColumnName: "PredictedSalary");

    after train above data how we can save the result set in a file.

    1. How to train the model over and over again with new data?

    2. How to get multiple Prediction (“Labels”) as an e.g, given by customer name or ID need to get predicted product list.

    3. How to select best training algorithm according to business problem?

    Gehan Fernando Snr. Engineer Technology. AKLO Information Technologies (Pvt) Ltd. #58, 42nd Lane, 5th Floor, Wellawatta, Colombo 06, Sri Lanka. Phone: +94 117 520000 | Mobile: +94 772 269625

    Sunday, October 13, 2019 3:56 AM

All replies

  • Hi 

    Thank you for posting here.

    Since this thread is related to ML.net, it is a third-party product. I don't suggest that you could post it here.

    The Visual C# forum discusses and asks questions about the C# programming language, IDE, libraries, samples, and tools.

    Best Regards,


    MSDN Community Support
    Please remember to click "Mark as Answer" the responses that resolved your issue, and to click "Unmark as Answer" if not. This can be beneficial to other community members reading this thread. If you have any compliments or complaints to MSDN Support, feel free to contact MSDNFSF@microsoft.com.

    Monday, October 14, 2019 3:16 AM