AI and ML of NTMA in here
👉 A new version of Model AI For IDS is available
- Model Non Label using LSTM --> File
- That will be base on the existing LSTM implementation, just renew some kind for like stuff work (CCIDS2017-Data and more ...)
- LSTM implementation in OC model - easier to read and understand
- Model with Label using Kmean --> File analysis and Kmeans for queue message
- That module is available in used in python3 but it need env file to connect with prometheus and kafka to trigger queue messages and process analytic and predict anomalies
- Require for run this model --> Huge resource will be use on init step for model 1-2Gb Rams (Need available) --> After that it will be stable and need resource when retrain model
- Just queue and wait to detection - Anything damage will be notify for your telegram bot
Design of KMeans IDS for Prediction Anomalies and Handling Firewall Passive
👉 A new version of Model AI For Forecasting metrics is available
- Model Non Label using LSTM --> File
- The model utilizes data retrieved from calling the Prometheus API to extract specific data. This data will be processed and stored in a file format at here!
- That module is available for use in Python 3, but it requires an env file to connect to Prometheus in order to access the API and collect data.
- Require for run this model --> Huge resource will be use on init step for model 1-2Gb Rams (Need available) --> After that it will be stable and need resource when retrain model
- After the model operates normally, when it predicts anomalies, the model will automatically enable autoscaling mode.
Design of Forecasting metrics and autoscaling
