Please enable JavaScript.
Coggle requires JavaScript to display documents.
Main Project - Coggle Diagram
Main Project
AI-Based Log Analysis
Project Objective
Analyse system and network logs
Detect anomalies and suspicious activity
Identify potential security threats
Support proactive monitoring and incident response
Data Collection
Kaggle log / intrusion-detection dataset
System logs
Network logs
Application / server logs
Data Preprocessing
Remove duplicates and unnecessary fields
Handle missing values
Parse log messages
Convert timestamps
Text normalization
Feature extraction
Exploratory Data Analysis (EDA)
Analyse log patterns
Normal vs abnormal activity
Event frequency
IP and user activity
Error and login trends
Visualise patterns over time
Feature Engineering
Failed-login count
Request frequency
Connection frequency
Time-based features
Categorical encoding
Machine Learning
Supervised Learning
Logistic Regression
Random Forest
XGBoost
Unsupervised Learning
Isolation Forest
One-Class SVM
Autoencoder (optional)
Model Evaluation
Accuracy
Precision
Recall
F1-score
Confusion Matrix
ROC-AUC
Anomaly Detection
Detect unusual behaviour
Score suspicious events
Identify potential attack patterns
Generate alerts
Explainable AI
SHAP feature importance
Explain anomaly predictions
Identify important log features
Improve model transparency
Dashboard & Deployment
Streamlit dashboard
Anomaly summary
Attack/event trends
Suspicious IPs
Model performance
Interactive filters
Tools & Technologies
Python
Pandas
NumPy
Scikit-learn
XGBoost
Matplotlib / Plotly
SHAP
Streamlit
Kaggle
PostgreSQL / MySQL
Final Outcomes
Reliable anomaly detection
Improved log monitoring
Faster incident identification
Data-driven security insights
Level 7 evaluation and model comparison