Self-introduction

My previous job was at SIer, and my name is @mizai, who has been involved in the development of web applications for about 13 years. I mainly used Java on the server side. (At the time of joining the company, there was a possibility of becoming a COBOL engineer ... (Shimijimi))

Most of the time I made business systems, and I went to various sites in various industries such as pharmaceutical companies, housing manufacturers, food manufacturers, and banks.

At the site of making a banking system, I was doing offshore (China) development, and I actually went to China for technical guidance. (It's a secret that I went to buy "Walking the Earth" as soon as I decided to go on a business trip to China. <-What are you going to do! W)

As a Java engineer, I would like to write down the results of doing machine learning with Python. (Of course, I have never used Python in my business, and I have little knowledge of Python.)

Background

Currently, I belong to a team that is in charge of system development for other companies, but since we have a large amount of data, we hope that the team that handles the data will call us at any time. I started basic learning to get ready.

This is the book I'm using. I chose it because I could do it while actually writing the code.

Environment

The first thing I was confused about was building a Python development environment. First, where to install various tools and libraries, the book showed how to specify the version.

What is Anaconda? ?? Well, yeah.

Reference: Article that writes about Anaconda (Is it convenient for machine learning?)

In the book, Python 3.6 is written, but 3.7 came out ?? Well, yeah. I like the latest and try to proceed with 3.7.

So, where to install the library Install variously while specifying the version with the commands conda and pip.

After all, what is the difference between this command? Well, yeah.

Really, in some library (I forgot what it was)

** Cannot install! I got an error like **. .. ..

Whaaaaaaaat! !! !! I looked it up and found that it wasn't Python 3.7. So, if I dare to downgrade (?) To Python 3.6, it succeeded.

Oh, it seems that the version is quite severe. .. .. That is my first impression.

[Reference] This is the screen of Anaconda

development of

The environment is ready and programming is ready! !! When it comes to that, I use something called "Jupyter Notebook" (hereinafter referred to as "Jupyter").

When you start Jupyter from Anaconda while following the instructions in the book,

Like this.

e? Somehow the browser started up? ??

Apparently, the style is to write the code in the column labeled In and execute it. Write the code and press "Shift + Enter" to execute it.

When you're advancing the basics of Python with Jupyter **what's this? Why is this a development environment? ?? Do I need to use this? ?? ** ** This is what I thought.

Yes, until the story of machine learning begins ...

Data visualization

When the basics of Python were over and the story of machine learning began ** Oh! This may be convenient! !! ** **

that's what I thought.

why! ??

If you write a script and execute it like this, you can visualize the data! !!

I see, it's a quick way to visualize data like this. ** It would be annoying to try it in Java. ** ** ** Python Convenient and interesting! ** ** I thought.

However, I wondered if it would be a little difficult to do large-scale development with Python. (Large-scale development referred to here: Creating a large business system in which various companies and people of various levels collaborate)

** 1. Experiment with data processing in Python ** ** 2. When making a product, write it in Java (a hard language like) **

I wonder if it should be used properly.

Even in fields I haven't experienced before, when I touched it even a little, I felt that the width and depth that I could imagine would increase at once.

The result of Java engineers learning machine learning in Python

Even if it's your first language, you can manage it!
After all, you should write the sample code yourself and experience it!
The language is the right person in the right place! You should learn if necessary!
AI looks fun ~ ☆

I thought. (Small feeling

** If you have any ideas for AI / IoT-like services, be prepared to create them! !! !! ** **

After finishing this book, I will study ** blockchain ** as well. (Oh! I confessed that the machine learning book wasn't over yet ... ORZ)

Finally

If you were thinking of summarizing a comparison between Java and Python, it was already summarized here. https://www.sejuku.net/blog/36782

end

The result of Java engineers learning machine learning in Python www