Where to start developing a neural network?

Question

Neural networks

Good evening toaster!
I seem to know C ++ normally, at an intermediate level, and I became interested in the topic of neural networks and machine learning. A search for this topic in Google did not bring the desired result - most of the articles deal with mathematics, complex formulas, or even biology at all (although, of course, I understand that this cannot be avoided, but I would like to find more practical information ). I'm thinking of creating a network for text recognition from pictures (as well as recognizing the font of this text) or captcha. Where can I find more or less organized information on this topic with specific code examples in order to somehow systematically develop in this area? Or where can you learn it? Is it realistic to learn something in this area on your own, or is it just a Google developer?

Answer 1

To begin with, I recommend this stack:
python->theano->keras
Under Windows, all this is already preinstalled and configured in Anaconda, it expands into 4 commands.

Answer 2

https://www.google.com/search?&q=python+neural+net...
iamtrask.github.io/2015/07/12/basic-python-network

Answer 3

Not the greatest experience, but if you really want to understand how to use neural networks, then it is better to use python. The ecosystem is well developed there, it is possible to use cuda. Have a look at keras and try to solve something on kaggle. Should be enough for the first time.
In the process, you usually want to spend less time on the technical features of the language. Any problems with retraining there will be enough for the eyes.

Answer 4

I highly recommend the course:
https://www.coursera.org/learn/machine-learning/ho...
It explains everything in detail from the very basics. You can just look separately about neural networks, but everything is desirable in order

Answer 5

Examples of neural networks in Matlab.