Confusions about hopfield neural network solved example

[shivajikobardan]

I am from Nepal and instead of asking us code, these are types of questions that are asked in our country examination system.
IDK what's their purpose. But some of them are badly hard as there are not many examples about it in textbook(In Nepal we can't get the textbook reference that the syllabus is made upon, we can of course get international writers textbook but our syllabus and exam paper is based on Indian author textbooks and we don't get that here in Nepal)
So I was studying about hopfield network and got badly confused. I have listed my confusions with annonations. It is not like I don't understand anything, I do understand the gist, but I am not very clear about each and every concepts.
And I didn't even find proper resources to study this topics in reference to our exam in internet as well.



Source-:

 

[Future Bruno]

1. What is auto-associative memory?Auto-associative memory is a type of neural network that is used to store and recall patterns. It is composed of neurons that are connected in a way that allows them to “remember” patterns that have been previously input into the network. By using the weights between the neurons, it is possible to reconstruct the original pattern from a set of partial inputs. It is typically used for pattern recognition, image processing, and speech recognition.2. How is Hopfield network different from other auto-associative memories?Hopfield networks differ from other auto-associative memories in that they are recurrent networks, meaning that the output of one neuron can be connected back to the input of another neuron. This allows the network to remember patterns more accurately, as the connections between neurons reinforce the patterns. Additionally, Hopfield networks are able to store multiple patterns simultaneously, while other auto-associative memories are limited to storing only one pattern at a time. Finally, Hopfield networks use an energy function to judge when a pattern has been correctly recalled, whereas other auto-associative memories typically rely on threshold values.