8a8de12fb3
* added training/batch size as cmdargs Signed-off-by: Toni Uhlig <matzeton@googlemail.com> |
||
|---|---|---|
| .. | ||
| c-analysed | ||
| c-captured | ||
| c-collectd | ||
| c-json-stdout | ||
| c-simple | ||
| cxx-graph@3491eafcd1 | ||
| js-rt-analyzer@87cb7a0af5 | ||
| js-rt-analyzer-frontend@48860b23c9 | ||
| py-flow-dashboard | ||
| py-flow-info | ||
| py-flow-muliprocess | ||
| py-json-stdout | ||
| py-machine-learning | ||
| py-schema-validation | ||
| py-semantic-validation | ||
| ndpid_install_and_run.gif | ||
| README.md | ||
examples
Some ready-2-use/ready-2-extend examples/utils. All examples are prefixed with their used LANG.
c-analysed
A feature extractor useful for ML/DL use cases.
It generates CSV files from flow "analyse" events.
Used also by tests/run_tests.sh if available.
c-captured
A capture daemon suitable for low-resource devices. It saves flows that were guessed/undetected/risky/midstream to a PCAP file for manual analysis.
c-collectd
A collecd-exec compatible middleware that gathers statistic values from nDPId.
c-json-stdout
Tiny nDPId json dumper. Does not provide any useful funcationality besides dumping parsed JSON objects.
c-simple
Integration example that verifies flow timeouts on SIGUSR1.
cxx-graph
A standalone GLFW/OpenGL application that draws statistical data using ImWeb/ImPlot/ImGui.
js-rt-analyzer
js-rt-analyzer-frontend
py-flow-info
Console friendly, colorful, prettyfied event printer.
Required by tests/run_tests.sh
py-machine-learning
Contains:
- Classification via Random Forests and SciLearn
- Anomaly Detection via Autoencoder and Keras (Work-In-Progress!)
Use sklearn together with CSVs created with c-analysed to train and predict DPI detections.
Try it with: ./examples/py-machine-learning/sklearn_random_forest.py --csv ./ndpi-analysed.csv --proto-class tls.youtube --proto-class tls.github --proto-class tls.spotify --proto-class tls.facebook --proto-class tls.instagram --proto-class tls.doh_dot --proto-class quic --proto-class icmp
This way you should get 9 different classification classes. You may notice that some classes e.g. TLS protocol classifications may have a higher false-negative rate. Unfortunately, I can not provide any datasets due to some privacy concerns.
But you can use a pre-trained model with --load-model.
py-flow-dashboard
A realtime web based graph using Plotly/Dash. Probably the most informative example.
py-flow-multiprocess
Simple Python Multiprocess example spawning two worker processes, one connecting to nDPIsrvd and one printing flow id's to STDOUT.
py-json-stdout
Dump received and parsed JSON objects.
py-schema-validation
Validate nDPId JSON strings against pre-defined JSON schema's.
See schema/.
Required by tests/run_tests.sh
py-semantic-validation
Validate nDPId JSON strings against internal event semantics.
Required by tests/run_tests.sh