1: \begin{abstract}

2: We introduce probabilistic neural networks that describe unsupervised synchronous learning on an atomic Hardy space and space of bounded real analytic functions, respectively. For a stationary ergodic vector process, we prove that  the probabilistic neural network yields a unique collection of neurons in global optimization without initialization and back-propagation. 

3: During learning, we show that all neurons communicate with each other, in the sense of linear combinations, until the learning is finished.

4: Also, we give convergence results for the stability of neurons, estimation methods, and topological statistics 

5: to appreciate unsupervised estimation

6: of a probabilistic neural network.

7: As application, we attach numerical experiments on samples

8: drawn by a standing wave.

9: %examples using the Koopman mode decomposition.

10: \end{abstract}

11: