Neural network class (AML part 1)

    A neural network is a complex data classifier. Special libraries are used to transfer data, such as sound or image, to the neural network. These libraries provide a digital representation of the object. Weights in the network are represented by numbers from 1 * 10-16 to 0.9.

    For images, I am using imageio for python. This library is there In the book of Tariq Rashid “Make Your Own Neural Network” there is a detailed description of how to create a neural network. However, there is no description of how to save weights after training the network.

    It would be very convenient to use a ready-made and trained neural network for work. Better yet, automatically start the learning process and get such a trendy tool today as automatic machine learning or AML.

    I will use the Python 3 programming language and its libraries for this. I’ll take the code from the book as an example, but I’ll modify it to suit my needs. I will create a neural network class that I will use in the future. I need to store the weights of a neural network in a database.

Link to source code from book