Sudipta Mukherjee - ML.NET Revealed: Simple Tools for Applying Machine Learning to Your Applications

Any source code or other supplementary material referenced by the author in this book is available to readers on GitHub via the books product page, located at www.apress.com/9781484265420 . For more detailed information, please visit http://www.apress.com/source-code .

Thanks for picking this book. This will introduce you to the wonderful world of machine learning via Microsofts open source cross-platform framework ML.NET.

That means if you master this framework, you can write machine learning (a.k.a. ML) applications or applications that use ML and run it on all platforms (Windows, Linux, MacOS).

Here is a brief introduction to the chapters.

Chapter : Meet ML.NET (Nothing is magical, but a few things seem so)

This chapter introduces you to the ML.NET framework and gives a very brief overview of tasks that are possible via ML.NET.

Chapter : The Pipeline (Great Machine Learning requires great plumbing)

This chapter introduces you to the plumbing that needs to happen in order for your ML tasks to be successful.

Chapter : Handling Data (Cleansing is engineering)

Data come in different formats and mostly are messy when they are onboarded in a system. This chapter shows how to clean data using several transformations offered by ML.NET.

Chapter : Regressions (How much will our dream home cost?)

This chapter shows how to use regression algorithms to predict prices of things in the future.

Chapter : Classifications (Helping computers tell chalk and cheese apart)

Classifying one object from another (a.k.a. binary classification) and classifying many objects in different categories (a.k.a. multiclass classification) are two classic ML tasks that are solved using ML.NET in this chapter.

Chapter : Clustering (Birds of a feather flock together)

Grouping things automatically into different groups is called clustering, and this is a classic unsupervised learning algorithm. This chapter shows how to solve these problems using ML.NET framework.

Chapter : Sentiment Analysis (Are you happy or not, that's the question!)

Automatically detecting polarity (positive or negative) from phrases is really important business and is an active research area. This chapter shows how to do sentiment analysis using ML.NET and some other techniques that are yet to appear in ML.NET but will sure do soon.

Chapter : Product Recommendation (You might be interested in this movie)

Product recommendation boosts product sales, and this chapter shows how you can use popular techniques like collaborative filtering and matrix factorization using ML.NET for product recommendations.

Chapter : Anomaly Detection (That doesn't look normal. Does it?)

Detecting odd ones from a pool of products is key to the success of a manufacturing business at this time because it is inhuman to expect that human employees can monitor everything. Thats where anomaly detection comes in to help. This chapter is dedicated to those algorithms and how to do those using ML.NET.

Chapter : Object Detection (Can you spot the cat in the photo?)

Detecting objects, faces from a photo or video frame is all the trend these days and has many applications. This chapter shows how to use ML.NET to do object detection using deep learning via ML.NET and ONNX.

A book like this one is not the fruit of a single persons labor. This is a team effort although that doesnt quite show up on the surface, where the name on the books jacket reflects that of the author. So I take this opportunity to express my gratitude for people without whom the book wouldnt have hit the press at all.

Writing is hard. If you have written anything significant, you know that you shall suffer from writers block, feeling completely blank in your head about what to write while funnily enough have a clear understanding of what would the pretty picture look like when the writing gets completed. In these situations, you need people who understand that situation and keep backing you up and wait for the content. I am not only fortunate but would say rather blessed to work with two individuals with a great deal of patience and grit. They are my acquisitions and coordinating editors Joan Murray and Jill Balzano. They constantly waited for my contents to appear and when I failed waited again. At moments I felt like I am walking on a slippery slope and wouldnt probably make it in any time at all, let alone soon. But miracles do happen when you have such wonderful people around you. Thanks a lot Joan and Jill. It was quite a ride and am surely looking forward to more in the future. Hopefully I shall disappoint you a little less in the future.

The next person I am indebted to is Olia Gavrysh. She is a program manager in the .NET team and previously managed the ML.NET team. She agreed to review the text when I approached her. This is very kind of her and she was very fast and accurate in providing eye-opening feedback that really improved the quality of the book and expanded my knowledge as well. When someone like herself, who has spent quite a lot of those initial days with the ML.NET team, comes forward and reviews the text, and when she approves what I have to say about ML.NET, it means a lot to me. I cant thank you enough Olia for doing this for the book. I seriously hope that we remain connected for future projects!