I need cybersecurity datasets for training machine learning algorithms, they could be for:
- intrusion detection
- DDoS detection
- Malware detection
- etc.
Check your local government website, but from that source you will get less number of data. I found these datasets from https://www.kaggle.com/ and have used in several projects.
You can use the UNSW-NB15 dataset.
https://github.com/InitRoot/UNSW_NB15
Dear Konstantin Dimitrov ,
Take a look at: https://analyticsindiamag.com/top-8-cybersecurity-datasets-for-your-next-machine-learning-project/
1) Digital Image Forensics: Image Manipulation or forgery detection.
2) Deep fakes: There are lots of datasets in kaggle and github
Top 10 Datasets For Cybersecurity Projects.
- MAWILab
- Malware Training Sets
- Comprehensive, Multi-Source Cyber-Security Events
- Malicious URLs
- ADFA Intrusion Detection Datasets
- Unified Host and Network Data Set
- User-Computer Authentication Associations in Time
- CTU-13 Dataset
- Aposemat IoT-23
- EMBER
Take a look at : https://analyticsindiamag.com/top-10-datasets-for-cybersecurity-projects/
If still interested, we recently developed a new dataset called USB-IDS-1, named after our university. The dataset provides several DoS attacks (hulk, slowloris, TCP flood, ...) tested against different state-of-practice web-server defense modules, such as mod_evasive and mod_reqtimeout (and, of course, the baseline 'no defense' case for each attack). We collect network flows for each attack, and all the records are labeled (annotated). Data can be used for research purposes by mentioning the source and referencing the related paper. Website: http://idsdata.ding.unisannio.it/
- Badges
- Science topic