← Back to all posts
11
Recognizing Handwritten Digits From the MNIST dataset!

This a neural network that can recognize handwritten digits! It uses the MNIST dataset => http://yann.lecun.com/exdb/mnist/. You can try the neural network out here => https://neuralnetwork.ieatpython.repl.co/ . It is not the most accurate at this moment and I plan in the future to make it so it can learn from the data!

Make sure to leave an upvote if the neural network got your number right! Also make sure to leave an upvote if it did not get it right because soon I will make it learn from its mistakes!

Here is a gif of it in action!:

Finally upvote if you enjoyed this or you think this was cool.

Commentshotnewtop
1

HAHA! this is awesome, nice work!

1

This is awesome! How did you learn how to do neural nets?

2

@amasad I started a while ago when I found this video on neural networks. It was amazing. https://www.youtube.com/watch?v=aircAruvnKk&t=152s
It taught about how neural nets work. In that video he used recognizing handwritten digits as an example. That was what made me want to make a neural network to recognize digits.

I chose HTML CSS and JS as a choice to make the neural network. In JS there is library called brain.js which is a JS library for neural networks. From there I really just watched a lot of brain.js videos and also did some reading on it and I got to the piece I have today!

1

I like the idea but it doesn't work for me.

1

@IEATPYTHON Seems to be my browser settings. It doesn't work with Firefox but Chrome is fine.

Edit: Great job! Works nearly perfect.