steps: prepare data set → write the index of the data set into 'csv4trainin.txt' (and csv4testing.txt) → training DNN models
''' the first step: prepare data set baiduSpider, developed on Python3
run in terminal as follows:
python spiderBaidu.py
You can substitute 'Your key words' at the end of the spiderBaidu.py
'''
''' the second step: write the index of the data set into csv.txt, which is a standard data format in TF.
You may refer to https://www.tensorflow.org/versions/r0.11/how_tos/reading_data/
(The csv.txt provided here is just an example. You might need to implement this step by yourself) '''
''' the final step: training DNN models: run in terminal as follows:
python DNN.py
the prediction results and training time are simply stored in thePredResults.txt '''