Table 7 Parameters of the machine-learning regression models used for training
From: A Python toolbox for neural circuit parameter inference
Model | Parameter | Value | |
|---|---|---|---|
MLPRegressor | hidden_layer_sizes | from (25, 25) to (50, 50) for catch22, from (2, 2) to (4, 4) for 1/f slope | |
max_iter | 100 | ||
tol | 0.1 | ||
n_iter_no_change | 5 | ||
Ridge | alpha | [0.01, 0.1, 1, 10, 100] | |
NPE | density_estimator | hidden_features | 10 for catch22, 2 for 1/f slope |
model | maf | ||
num_transforms | 2 | ||
