MNIST – the totally non-clickbait beginnings

Hello World

As every proper blog, especially written by a programmer, there’s a need to write a “Hello World” post.

I don’t want to be special in this case, so here you’re!

I’d like to present you MNIST Hello World.

Cool.

But what’s MNIST?

Hint: let me use another sophisticated acronym – WYSIWYG (you should have heard about it before)

Look at the picture again. What can you see?

.

.

.

.

.

.

.

Handwritten digits?

Yep!

Being more accurate, MNIST is a database of handwritten digits very often used as a training set in the field of machine learning. When you’re, like me, a beginner in ML, it’s likely you will use it to write your very first program which would be able to recognise, unknown yet, handwritten digits.

But we won’t use this on this blog anyway. I just didn’t have a better idea for the “Hello World”. And this MNIST printscreen looks nice, especially with that dark theme. Doesn’t it? 😀

I’m excited about machine learning as much as about creating the music. I’m going to combine these two fields together and make playing the guitar equally important as coding. Even more, I’m going to do these two activities almost simultaneously. Feeding a neural network with sounds produced by guitar is sexy – at least for me.

But before it all will happen, I need to learn. I need to learn a lot about machine learning. Ok, let’s name things what they are. My knowledge about this field at the moment is close to (zero). I need to learn machine learning from scratch.

I have a goal to understand and use Magenta which is a build on top of Google’s Tensorflow. I’m going to get a good course about machine learning and publish the progress of studying it on the blog. Then, I should be aware enough to start playing (a little) with Magenta.

Here is my GitHub. I’ll commit all the stuff I found useful in the way of learning, and finally and hopefully, in a few weeks, there should appear a bigger “project”. At the moment, I’m thinking about AI-powered ambient generator running in a browser.

What’s actually going to happen? Nobody knows. The statistics would say I’ll fail. At this point, I could write at least one good excuse why I might fail. But this post isn’t about excuses. Maybe the next one?

It’s time to begin a not short, not easy and not boring journey which is, I have no doubt, writing a blog about machine learning.

Welcome you, my dear reader. I’m glad you’re here.

Stay tuned!

PS. I had a problem to configure widget for a fresh facebook page, so I share it with you in the classic way.

Want to follow my machine learning story?

2 comments

Leave a Reply

Your email address will not be published. Required fields are marked. *

Related articles