Machine learning is a fascinating world and, to many, a complicated one. As .NET developers, we definitely see the benefit in training our data but between the learning curve and using other languages like Python for machine learning—a language .NET devs might not be familiar with—ML is often sent to a developer’s “I should look into that sometime” queue.
That changed in 2018, when Microsoft launched ML.NET—a free, open source, x-plat machine learning framework for .NET. With ML.NET, you can use your favorite languages like C## or F## to work with your custom machine learning models. The idea is to meet you where you are and make ML more accessible.
There’s no one better to talk to about this than Luis Quintanilla. Luis has been with ML.NET since the beginning and was eventually scooped up by Microsoft to work on the docs for ML.NET. Luis had so much great stuff to share that we’ll split this interview up into two parts. Last time, we talked about his path to Microsoft, the value of ML.NET, and how to get started. Today, we’re talking about using ML.NET over something like Azure Cognitive Services, use cases for ML.NET, and more.
Where is the dividing line for when I should use machine learning, or use Azure Cognitive Services?Permalink
This is a really tough one to answer because there’s so many ways you can make the comparison. Both are great products and in some areas, the lines can blur. Here are a few of them that I think might be helpful.
Custom Training vs ConsumptionPermalink
If you’re looking to add machine learning into your application to solve a fairly generic problem, such as language translation or identifying popular landmarks, Azure Cognitive Services is an excellent option. The only knowledge you need to have is how to call an API over HTTP. Being able to work via HTTP also provides you with flexibility over what you use to make the requests to the API. If you want to run your machine learning workflow with Azure Cognitive Services via cURL as part of a background Cron job, that’s perfectly acceptable.
Azure Cognitive Services provides a set of robust, state-of-the-art, pretrained models for a wide variety of scenarios. However, there’s always edge cases. Suppose you have a set of documents that you want to classify and the terminology in your industry is rare or niche. In that scenario, the performance of Azure Cognitive Services may vary because the pretrained models most likely have never encountered some of your industry terms. At that point, training your own model may be a better option, which for this particular scenario, Azure Cognitive Services does not allow you to do.
#[object object] #[object object]
