The phrase "at hidden layers" is correct and usable in written English.
It can be used in contexts related to machine learning, neural networks, or any discussion involving layers of abstraction or complexity that are not immediately visible.
Example: "The model's performance can be significantly improved by analyzing the features extracted at hidden layers."
Alternatives: "within concealed layers" or "in the underlying layers".
The ANN based on the Levenberg Marquardt algorithm (LMA) combined of linear transfer function at output layer and tangent sigmoid transfer function at hidden layer with 20 hidden neurons supply best operation conditions for good prediction of adsorption data.
After back-propagation (BP) training combined with principal component analysis, the three-layer ANN model with a tangent sigmoid transfer function at hidden layer with 11 neurons and a linear transfer function at output layer was obtained.
It was found that using ANN model with 12 neurons at hidden layer for all three ions, the R2 of 0.989 and MSE of 0.993 for Hg2+ ion were obtained.
The ANN based on the Levenberg Marquardt algorithm (LMA) composed of linear transfer function (purelin) at output layer and tangent sigmoid transfer function (tansig) at hidden layer with 5, 8 and 9 neurons for MG, DB and MB, respectively, give best operation conditions for good prediction of behavior adsorption system.
A tan-sigmoid transfer function (tansig) at hidden layer and a linear transfer function (purelin) at output layer were applied.
Also, let the corresponding hidden representation of d (x,y) at hidden layer be (a^{(x,y)}in mathbb {R}^{K}) and the reconstructed output be ({tilde {d}}^{(x,y)}in mathbb {R}^{n}).
The ANN model established in this study comprised three layers as shown in Fig. 7a consisting of a hyperbolic tangent transfer function at hidden layer and a linear transfer function at output layer.
The network is tested with various numbers of neurons from 1 to 12 for finding the optimum number of neurons at hidden layer by finding the mean squared error (MSE).
According to Fig. 6a, b, the optimum number of neurons at hidden layer is equal to 5 as the best case with the minimum value of MSE (8.268E−06 and 8.815E−06) for both As III) and As V).
