Inspired by - https://github.com/markbrutx/task
The goal of this project is to develop and train the MDN-RNN (Mixture Density Network - Recurrent Neural Network) model for analyzing the network traffic and indicate the suspicious entries. The inputs are the log files and the ourputs are entries that model label as anomalies.
The project generates an HTML file containing a full report of the model predictions as well as short recomendations on how that incidents may be avoided in future.
It is our first project with machine learning so it may not be perfect, but we've tried our best to make it as good as possible.
This project enables the prediction of cybersecurity incidents using the MDN-RNN model, which helps to proactively assess the probabilities of various threats and develop strategies to prevent them, thereby improving information security and system resilience to potential attacks.