Where to start learning neural networks in python?

Question

Python

Please tell me literature, courses, maybe libraries in python. From what I tried to read a lot of mathematics, but without practice it is not clear

Answer 1

A bunch of math - that's the way it is. Machine learning is essentially a large number of matrices, vectors, derivatives, etc., in other words, it is more of a mathematical discipline than a computer science discipline.
Yandex lectures
Ski-kit examples Ski-
kit
docks Pybrain docks
Machine learning in python book
edx
courses Stanford university courses are the same as on coursera udacity
courses
The most modern library for machine learning from google -
tensor flow

Answer 2

Library for practical tasks: scikit-learn.org
But it's better to start with the theory in order to understand what's what.
I recommend coursera - machine learning: https://www.coursera.org/learn/machine-learning

Answer 3

There is a course on stepic.ru - Neural networks , try starting with it. I myself have not watched it yet, but judging by the table of contents, there is practice there.

Answer 4

Keras, Teano, Tensorflow (more complicated and easier at the same time)
there are examples - lessons, even in Russian,
neural networks cannot be studied without knowing mathematics,
but you can use it quite successfully

Answer 5

First, read articles on how to implement neural networks and their learning algorithms, for example, a series of articles on Habré: https://habrahabr.ru/post/312450/
Only then implement them programmatically, I recommend using TensorFlow for this:
TensorFlow API: https:// www.tensorflow.org/api_docs/
Article on Habré: https://habrahabr.ru/post/305578/