Order of setting up project:

Step1:-

1. install required libraries using command: "pip install -r requirements.txt"
2. dowload dataset, make a folder dataset and make three subfolders: Bot-IoT,UNSW-NB15,cleaned_dataset
3. download link: usnw_dataset_link:https://cloudstor.aarnet.edu.au/plus/apps/onlyoffice/s/2DhnLGDdEECo4ys?fileId=206777051 and https://cloudstor.aarnet.edu.au/plus/apps/onlyoffice/s/2DhnLGDdEECo4ys?fileId=206779316 and put it inside dataset/UNSW-NB15
4. download bot_iot data_set: https://cloudstor.aarnet.edu.au/plus/s/umT99TnxvbpkkoE?path=%2FCSV%2FTraning%20and%20Testing%20Tets%20(5%25%20of%20the%20entier%20dataset)%2FAll%20features and put it inside dataset/Bot-Iot

Step2:-

1. Run data_cleaning_unsw.ipynb and data_cleaning_bot_iot.ipynb, to clean dataset
2. Copy the cleaned_dataset folder into normal_ddos_dos_classification
3. Run the training.ipynb, to train the model
4. Run the test.ipynb, to test the model

Common features used:

Spkts in bot ~ Spkts in unsw (Source-to-destination packet count) Dpkts in bot ~ Dpkts in unsw (Destination-to-source packet count) Sbytes in bot ~ sbytes in unsw (Source-to-destination byte count) Dbytes in bot ~ dbytes in unsw (Destination-to-source byte count) Dur in bot ~ dur in unsw (record total duration) Proto in bot ~ proto in unsw (transaction protocol) attack in bot ~ label in unsw (traffic: 0=normal traffic,1=attack traffic) stime in bot ~ Stime in unsw (record start time) ltime in bot ~ ltime in unsw (record last time) sport in bot ~ sport in unsw (source port number) dport in iot ~ dsport in unsw (destination port number) saddr in iot ~ srcip in unsw (source ip address) daddr in iot ~ dstip in unsw (destination ip address)

Ouput label: category in bot: (Dos/Ddos) category in unsw: (normal)