For the supplementary files with prefix:
1. Prefix "pin-" and "fix-" indicate that this file is for pinned-support/fixed-support shells;
2. Files "ini.txt" and "target.txt" contains the input and output training data;
3. Files "ini1.txt" and "target1.txt" contains the input and output testing data;
4. Function "myNeuralNetworkFunction.m" contains the trained parameters for the artificial neural networks (ANN), which can be directly used for prediction;
Note that when using the aforementioned data/code, the prefix should be removed.

For the supplementary files without prefix:
1. Function "rsuqare.m" is used for calculating the R-squared value of the prediction output against the precise data;
2. Run code "ann.m", and one can get the error distribution of the ANN prediction output against the precise data;
3. Run code "svr1.m", and one can get the error distribution of the support vector regression (SVR) using the Gaussian kernel prediction output against the precise data;
4. Run code "svr2.m", and one can get the error distribution of the SVR using the linear kernel prediction output against the precise data.

In order to use the models, one can prepare a file X.txt containing the input parameters, and:
1. Add the following code to the end of "ann.m":
   load X.txt
   Y = myNeuralNetworkFunction(X)
2. Add the following code to the end of "svr1.m" or "svr2.m":
   load X.txt
   Y = predict(Mdl,X)

Please contact the authors if you have any problem using the data/codes.