Introducing Microsoft machine learning library for .NET (ML.NET)

Possibly save 1 hour of your time: This is a quick introduction to Microsoft machine learning library ML.NET.

Microsoft

• https://www.microsoft.com/net/learn/apps/machine-learning-and-ai/ml-dotnet

Github

• https://github.com/dotnet/machinelearning
• Roadmap – https://github.com/dotnet/machinelearning/blob/master/ROADMAP.md

The library is only supported since .NET Core 2.0.

To start, you can create a .NET Core 2.0 console app. Then you can use nuget to find package Microsoft.ML and install it to your project.

Here are some features of the library:

 

Package is not compatible with Package supports netstandard2.0

Possibly save 1 hour of your time: When you try to install a nuget package with a dependency on .NET Standard 2.0, it gives you this error.

.NET Standard 2.0 is supported by .NET Core 2.

You can right click your project >> select properties. Check to see what .NET Core you are using.

If you don’t have .NET Core 2.0, the easiest way is to download latest is to go into Visual Studio > Tools > Extensions and Updates > Under Updates > Product Updates > you should be able to see Visual Studio Update.

You can also download SDK and Runtime here https://www.microsoft.com/net/download/visual-studio-sdks

For more information on .NET Standard 2.0, please see https://blogs.msdn.microsoft.com/dotnet/2017/08/14/announcing-net-standard-2-0/

 

 

LINQ query should not include your class projection to improve performance

Possibly save 1 hour of your time: I have found that performance of your LINQ query could be impact especially for large datasets when you project your results directly to a class object. It is better to project to .NET object after the sql query results came back first.

var results = (from a in context.TableA
                    join b in context.TableB on a.ID equals b.ID
                    select new YourClassName
                    {
                        Year = a.Year,
                        Name = b.Name
                    });
var list = results.ToList();

The code above is much slower than the code below for large datasets.

var results = (from a in context.TableA
                    join b in context.TableB on a.ID equals b.ID
                    select new
                    {
                        Year = a.Year,
                        Name = b.Name
                    });
var list = results.ToList();

var returnList = list.Select(x => new YourClassName
                {
                    Year = x.Year,
                    Name = x.Name
                }).ToList();