In order to get started with neural networks please maintain the given flow. First, you need to learn how neural networks work. Implement a simple code for neural networks.

First, go through these videos. This is the most intuitive explanation you will ever get of neural networks.

  • https://www.youtube.com/watch?v=aircAruvnKk&t=958s
  • https://www.youtube.com/watch?v=IHZwWFHWa-w&t=1s
  • https://www.youtube.com/watch?v=Ilg3gGewQ5U&t=9s
  • https://www.youtube.com/watch?v=tIeHLnjs5U8&t=86s

After this, I am guessing you will get a basic grasp of neural networks. Then you can go through these tutorials.

  • https://towardsdatascience.com/first-neural-network-for-beginners-explained-with-code-4cfd37e06eaf
  • https://mlfromscratch.com/neural-networks-explained/#/
  • https://blog.goodaudience.com/artificial-neural-networks-explained-436fcf36e75

Now you might have a complete grasp of neural networks. You need to learn Tensorflow/Keras. Here are some of the resources for learning TensorFlow. The best place to learn TensorFlow is their official implementation

  • https://www.tensorflow.org/tutorials/quickstart/beginner
  • https://youtu.be/tPYj3fFJGjk (A Long 7 Hour workshop on TensorFlow, very helpful)
  • https://github.com/aymericdamien/TensorFlow-Examples (Some jupyter notebooks for TensorFlow tutorials)

As now you have the basic understanding of neural networks. Time to learn Convolutional neural network. Go through the following resources.

  • https://ujjwalkarn.me/2016/08/11/intuitive-explanation-convnets/
  • https://medium.com/@RaghavPrabhu/understanding-of-convolutional-neural-network-cnn-deep-learning-99760835f148

There are many available resources on the internet. You can take a look at them too.

Now as you have a sufficient amount of ideas CNN. Time to implement this.

  • https://www.youtube.com/watch?v=umGJ30-15_A (CNN Tutorial with tensorflow implementation)
  • https://www.tensorflow.org/tutorials/images/cnn
  • https://towardsdatascience.com/tensorflow-2-0-create-and-train-a-vanilla-cnn-on-google-colab-c7a0ac86d61b

After that, now you have to go for the hardest part. Learning Recurrent Neural Networks is the most difficult one. But here are the most useful resources for these.

  • https://www.youtube.com/watch?v=UNmqTiOnRfg
  • https://www.youtube.com/watch?v=y7qrilE-Zlc (Tutorial with tensorflow)
  • https://colah.github.io/posts/2015-08-Understanding-LSTMs/

Lets implement RNN with tensorflow

  • https://www.tensorflow.org/guide/keras/rnn

Now some auxiliary resources.

How to use Google Colab:

  • https://www.youtube.com/watch?v=inN8seMm7UI (Basic Intro)
  • https://towardsdatascience.com/getting-started-with-google-colab-f2fff97f594c

How to setup CUDA in ubuntu:

  • https://www.pugetsystems.com/labs/hpc/How-to-install-CUDA-9-2-on-Ubuntu-18-04-1184/

Deep Learning in a Nutshell

  • https://medium.com/datadriveninvestor/deep-learning-in-a-nutshell-a600a27ab371

Image Pre processing for neural networks

  • https://towardsdatascience.com/image-pre-processing-c1aec0be3edf