Fig. 3: Results for generation when source and target are the same region.

a Both training data and prediction, generation, data are prepared from the same region in this evaluation. b In the case of example (a), both training data (first 17 minutes) and test data (second 17 minutes) of the same time series are obtained from area A. c This panel shows the result of predicting the firing rate in this case, where the x axis is the firing rate in the original data and the y axis is the firing rate in the data generated by training Multilayer LSTM. d This panel shows the result of predicting synchronization score. Again, the x axis is the synchronization score in the original data, and the y axis is the synchronization score in the generated data. This data was expressed in r-θ rotational coordinates, and a histogram of the number of data in 0-π/2 with respect to θ, or in the first quadrant, was drawn in e. Finally, f is the histogram of the number of data in π-3/2π with respect to θ, or the third quadrant. In particular, if the output is coming from inhibitory cells, it is distributed in the third quadrant. The sharpness of the peaks in these histograms (e, f) was evaluated by sharpness [Refer to the method section]. g Correlations between expected and true values of firing rates in the inhibitory neurons are plotted for every 16 regions. The two points for every group of regions correspond to the two datasets, and the line is the averaged value between the two datasets. The meaning of the points and lines is the same for h, j. h Correlations between expected and true values of firing rates in excitatory neurons are plotted in the same way as in g. i shows correlations between expected and true values of synchronization score in the first quadrant for every 16 regions. j shows correlations between expected and truth values of synchronization score in the third quadrant.