Hello! I have an array of data, it is updated: the last element is deleted and the new one is added (the values of the random number generator). The question is that the array "A" is constructed in such a way that the new element is the last A [n], and the old element is the first A [0]. It is updated from right to left. Each time on the new iteration, the old value of A [0] is deleted and the new value is appended to the end of A [n]. Then there is training and so on until the termination of the run-out. Question! How does the array interpret the neural network at the input: 1) it does not matter? Or 2) as a chronological sequence of the set of numbers from the old to the new element and in which direction from left to right \ right to left?. It is important that ANN on a set of numbers predicts the next number ... Rather it is necessary that the network understood where the old element is A [0], and where the new A [n] .. Perl Language. Library AI :: NNFlex
https://www.researchgate.net/post/why_the_prediction_or_the_output_of_neural_network_does_not_change_during_the_test_phase
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.4.8794&rep=rep1&type=pdf
https://www.researchgate.net/post/How_to_decide_the_number_of_hidden_layers_and_nodes_in_a_hidden_layer
https://arxiv.org/pdf/1705.04977
https://arxiv.org/pdf/1802.08760
https://www.civisanalytics.com/blog/interpreting-visualizing-neural-networks-text-processing/
ftp://ftp.sas.com/pub/neural/importance.html
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence