I gave the last exam, neural networks, 10 days ago. So it appeared the day after my living space: chaotic, as usual, life.
The fact that even in dynamical systems that describe the simplest neural network is provided for the assumption of an asymptotically chaotic behavior of many phenomena can be self-explanatory: it is inherent in the rationality, then there is history.
Just as I can feel like a physical system in a ring of a Markov chain: the present situation is not influenced by past history, but there is only a probability of transition to one of the possible next states.
Not so with Markov chains, which make the decisions: a decision simply because it can not be taken if two or more alternatives there is a cheaper, then the decision does not even arise. If the alternatives are equivalent to or are unable to establish an order of priority, the decision can not be taken and you risk ending up in a periodic motion from which it does not exit.
The hope is that your attractor leads to something good.
Happy May Day to all workers and aspiring
.
.