Detect Attacks:

A python package which detects network attacks includes:

• Collecting data from attacks
• Classifying data to predict the risks of the network attacks
• Warning users risks which could be a network attack.

Getting Started

Prerequisites

• These packages should be installed before using detect_attacks:

tensorflow	1.5.0
sklearn	0.19.1
keras	2.1.3
numpy	1.14.0
matplotlib	2.1.2
deepmg 0.5.9

• Please install if you do not have them

pip install matplotlib
pip install numpy

conda install scikit-learn
conda install -c conda-forge tensorflow 
conda install -c conda-forge keras

pip install Keras-Applications
pip install Keras-Preprocessing
pip install keras_sequential_ascii

pip install deepmg

Install or Download the package detect_attacks

pip install detect_attacks

Running Experiments

How to use detect_attacks

• Input:

  • mandatory: csv files containing data (*_x.csv) and labels (*_y.csv)
  • optional: if use external validation set: data (*_zx.csv) and labels (*_zy.csv)) put in data changable with parameters --orginal_data_folder).

  For examples, data1_x.csv and data1_y.csv for.

• Output:

  • results: performance/training/testing information of each fold and summary results put in [results/name_dataset_parameters_to_generate_image/] (results/) (changable with parameters --parent_folder_results), includes more than 5 files:

    • *file_sum.txt: parameters used to run, performance at each fold. The last rows show training/testing performance in ACC, AUC, execution time, and other metrics of the experiment. When the experiment finishes, a suffix "_ok" (changable with parameters --suff_fini) appended to the name of file marking that the experiment finishes.

    • *file_eachfold.txt (if --save_folds=y): results of each fold with accuracy, auc, mcc, loss of training and testing.

    • *file_mean_acc.txt (if --save_avg_run=y): if the experiment includes n runs repeated independently, so the file includes average performance on k-folds of each run measured by accuracy and time execution at training/testing of beginning, training/testing when finished.

    • *file_mean_auc.txt (if --save_avg_run=y): if the experiment includes n runs repeated independently, so the file includes average performance on k-folds of each run measured by AUC at training/testing of beginning, training/testing when finished.

    • If --save_para=y: configuration file to repeat the experiment

    • If use --save_w=y (save weights of trained networks) and/or --save_entire_w=y, --save_d=y, then 2 folders will be created:

      • results/name_dataset_parameters_to_generate_image/models/: includes *weightmodel*.json contains structure of the model *weightmodel*.h5 stores weights.

      • results/name_dataset_parameters_to_generate_image/details/*weight_*.txt: contains accuracy and loss of training and testing every epochs --save_d=y. If --save_rf=y, then we will have important scores generated from RFs for each run.

Some examples

db_name='data1';
folder_data='/Users/hainguyen//test/data/';
folder_res='/Users/hainguyen//test/results/';

python -m detect_attacks -i $db_name -r $folder_data --parent_folder_results $folder_res --model rf_model
python -m detect_attacks -i $db_name -r $folder_data --parent_folder_results $folder_res --model svm_model
python -m detect_attacks -i $db_name -r $folder_data --parent_folder_results $folder_res --model model_cnn1d
python -m detect_attacks -i $db_name -r $folder_data --parent_folder_results $folder_res --model model_mlp
python -m detect_attacks -i $db_name -r $folder_data --parent_folder_results $folder_res --model fc_model

Download data

Please use the link below to download data for the experiments

darpa29f

References: