If you decide to train the GPR by yourself,after the training process finishes you can visualize in Tensorboard how the variables and metrics chance by doing the following in the terminal:
tensorboard --logdir=./tensorboard/
Requirements
install:
-
nibabel
-
tensorflow
-
Create "logs" folder
-
Store the log files in that folder
Current Version:
White Matter BANC_2016
Linear kernel, non-ARD
Results on 200 testing subjects:
***RMSE at testing time is :7.20819736674 ***MAE at testing time is :5.64525260886
**** How to use scripts ***
-
use create_dataset.py it has the following arguments:
'--mask_path' help='the absolute path to the nifti file containing the mask you want to apply'
'--data_info_path' help='the absolute path to the csv file containing meta data about yur dataset, such as BANC_2016.csv'
'--folder_nifti_path' help='the absolute path to the folder which contains your nifti files'
'--output_folder' help='the absolute path of the folder where you want to store the array files created by this script'
'--output_X_name' help='the name of the .txt file where we store the array containing the input features'
'--output_Y_name' help=' the name of the .txt file where we store the column-vector containing the chrnological age of subjects'
-
use GPR.py to train a new model if you want;otherwise go to step 3 it has the following arguments:
'--input_feature_path' help='the absolute path of the file which contains your features dataset'
'--input_age_path' help='the absolute path of the file which contains your chrnonological age columns-vector'
'--num_iterations' default=1000 help='the number of iterations in the training process'
-
use GPR_Prediction.py to get predictions for new subjects; obv use step 1 again to create the array folders for your externat dataset it has the following arguments:
'--input_feature_path' help='the absolute path of the file which contains your features dataset'
'--input_age_path' help='the absolute path of the file which contains your chrnonological age columns-vector'