WebsiteFingerprinting

Use deep learning to perform website fingerprinting.

Setup

mkdir outputs data

mv full.zip ./data/full.zip

unzip data/full.zip ./data/full

conda env create -f environment.yml

If you are having trouble getting GPU support with TensorFlow, try

pip uninstall black

conda uninstall black

conda install tensorflow-gpu

Usage

conda activate WF_Assigment

Please the code and comments contained within the main clause in main.py carefully. Adjust as desired.

python main.py

Output

Structure


-- "outputs"

   -- {OPTIMIZER}
   
      -- {LEARNING_RATE}
      
         -- {"RNN|LSTM|GRU"}
         
            -- {HIDDEN_UNITS}
            
               -- {NUM_LAYERS}
               
                  -- {BIDIRECTIONAL}
                  
                     -- "models"
                     
                        -- "{EPOCH}.pt"
                        
                     -- reports
                     
                        -- "{EPOCH}.pt"
                        
                     -- report.json
                     
         -- "CNN"
         
            -- {FILTER_SIZE}
            
               -- {ACTIVATION_FUNCTION}
               
                  -- history.json
                  
                  -- report.json

Files

RNN/LSTM/GRU

file: outputs/{OPTIMIZER}/{RNN|LSTM|GRU}/{HIDDEN_UNITS}/{NUM_LAYERS}/{BIDIRECTIONAL}/models/0.pt

example: output_file_examples/1.pt

description: saved model at a particular epoch

file: outputs/{OPTIMIZER}/{RNN|LSTM|GRU}/{HIDDEN_UNITS}/{NUM_LAYERS}/{BIDIRECTIONAL}/reports/0.pt

example: output_file_examples/2.json

description: model evaluated on the validation set at a particular epoch

file: outputs/{OPTIMIZER}/{RNN|LSTM|GRU}/{HIDDEN_UNITS}/{NUM_LAYERS}/{BIDIRECTIONAL}/report.json

example: output_file_examples/3.json

description: best model (PATIENCE epochs before early stopping) evaluated on the test set

CNN

file: outputs/{OPTIMIZER}/CNN/{FILTER_SIZE}/{ACTIVATION_FUNCTION}/history.json

example: output_file_examples/4.json

description: model evaluated on the validation set at every epoch

file: outputs/{OPTIMIZER}/CNN/{FILTER_SIZE}/{ACTIVATION_FUNCTION}/report.json

example: output_file_examples/5.json