Let’s start by defining what data science is and isn’t.

I define data science as the discipline of preparing and analyzing large amounts of data to extract insights and determine future behavior through machine learning and predictive modeling.

In other words, data science is all about gathering insights from the large amounts of data organizations amass every day. In fact, this is part of why data science has experienced an increase in popularity in the past decade.

Over the past several decades, organizations have moved more of their applications to be hosted on cloud computing providers such as Microsoft Azure, Amazon Web Services (AWS), and Google Cloud. This shift to cloud hosting had several benefits, including the following:

• Ease in scaling web applications and databases as usage grows
• Ease in adding new capabilities such as cloud storage and machine learning
• Automated tools for collecting, storing, and...

In programming, when you want to illustrate a concept or explore new coding techniques, it’s normal to create a small application such as a console program and write the code you need there. This can be helpful for validating and communicating ideas, making small examples, reproducing errors, or building small demos.

Like programmers, data scientists also sometimes need to perform small experiments. While it’s entirely possible to perform data science experiments by creating a new Python or .NET program, a far more common approach for data scientists and data analysts is to create a notebook.

A notebook is a combination of documentation in the form of Markdown cells mixed together with code cells. This combination of code and documentation allows you to provide rich formatted documentation via Markdown while also providing live executable code in code cells.

For years, when people have talked about notebooks in data...

In our example in the previous section, we saw how code cells could be executed to produce a result.

This work is carried out by a kernel designed for the programming language you’re working with. In the earlier example, we saw a code cell written in Python. Since this is Python code, it uses the Python kernel.

In this way, the Jupyter Notebooks environment can support multiple languages, including Julia, Python, and R, as shown in Figure 1.8:

Figure 1.8 – Different kernels available to the notebook

When you execute a Python cell, the code is sent to the Python kernel, which interprets it, executes it, and produces a result as illustrated in Figure 1.9:

Figure 1.9 – Jupyter Notebooks executing the Python kernel

This extensible design allows Jupyter Notebooks to support additional languages by adding additional kernels for those languages.

This is exactly what Polyglot...

Polyglot Notebooks is a VS Code extension that integrates into Jupyter Notebooks to provide additional capabilities and language support through the kernel mechanism.

The goal of Polyglot Notebooks is to enrich the notebook experience by moving the supported languages beyond just the original languages. Polyglot Notebooks accomplishes this by adding its own kernel, called .NET Interactive, which supports the following languages:

• C#
• F#
• JavaScript
• Python
• R
• PowerShell
• HTML
• HTTP Requests
• Mermaid
• SQL
• KQL

The philosophy of Polyglot Notebooks is that you should be able to choose the languages and tools you want to use for the task you’re trying to accomplish.

In other words, Polyglot Notebooks wants to allow you to pull data into your notebook via SQL, KQL, or a REST request, and then use C#, F#, or JavaScript to manipulate it before allowing you to visualize it using the capabilities...

In this chapter, we discussed the high-level fields of data science, machine learning, and AI.

While we covered these earlier in more depth, here is a consolidated set of definitions:

• Artificial intelligence is the broadest field and revolves around emulating aspects of behaviors found in humans and animals.
• Data science is the discipline of preparing and analyzing large amounts of data to extract insights and determine future behavior through machine learning and predictive modeling.
• Machine learning is a broad field involving applying mathematics and statistics to solve data problems. Machine learning includes supervised learning, unsupervised learning, and semi-supervised learning including reinforcement learning.
• Supervised learning involves applying statistical and mathematical techniques to model trends and relationships found in datasets.

In this chapter, we also discussed the role of notebooks in data science for conducting iterative experiments...

See the following resources for more information on the topics covered in this chapter:

• Polyglot Notebooks and .NET Interactive history: https://devblogs.microsoft.com/dotnet/dotnet-interactive-notebooks-is-now-polyglot-notebooks/
• Machine learning tasks – ML.NET: https://learn.microsoft.com/en-us/dotnet/machine-learning/resources/tasks#clustering
• GitHub CodeSpaces: https://github.com/features/codespaces
• Project Jupyter: https://jupyter.org/
• AlphaGo: https://deepmind.google/technologies/alphago/
• A computer code that transformed science: https://cs.lbl.gov/news-media/news/2021/project-jupyter-a-computer-code-that-transformed-science/
• Anaconda: https://www.anaconda.com/download/
• Markdown cells: https://packt.link/U9OUX
• Data storage market size trends: https://www.fortunebusinessinsights.com/data-storage-market-102991
• Additional online resources: https://MattOnML.NET/Books/DataScience/introduction/