Prediction of Coal Fire Power Plant Pollutants Emission

[Dalpi]

Dear all,
My master thesis is about the prediction of gas emissions from a coal-fired thermal plant using neural network.
With the data from the coal-fired thermal plant and using Python I made the prediction for the emission of Nox, SO2 and CO. However, the problem encountered was to get a neural network with an error within an acceptable standard (low MSE)
I would like your help on the method of selecting the input variables in order to obtain a network with a small prediction error.
Researching I found something about Garson's method that uses the weight of connections. Does anyone know how I should proceed so that my network reports the weight of each connection? What lines of programming in Python should I include to generate this data?
Any other suggest to classifier the input variables?
Thank you.

[hussainmujtaba]

Here is an implmentation if you are using keras
The code below accesses the output of the first Keras hidden layer, which is the input for the hidden activation function (second Keras layer).

hidden_layers = keras.backend.function(
[model.layers[0].input],  # we will feed the function with the input of the first layer  
[model.layers[0].output,] # we want to get the output of the first layer
)
hidden_layers([x])

Running this code outputs the data below, which is the sum of each input for neuron 1; for neuron 2 it also the sum of inputs, but subtracted by(bias), as indicated by the first figure.

To get weights for next layer run the below code

hidden_layers = keras.backend.function(
[model.layers[0].input],  # we will feed the function with the input of the first layer
[model.layers[1].output,] # we want to get the output of the second layer
)
hidden_layers([x])

[Dalpi]

Thank you very much for your help.