A tool for the visual analysis of Search Trajectory Networks.
STN Analytics is the result of a multi-year research collaboration between the IIIA-CSIC and the University of Stirling. This tool implements and extends a novel methodology for understanding the behavior of complex optimization algorithms.
STN Analytics: A Comprehensive Tool for Dynamic and Multi-View Analysis
This upcoming paper formally presents the STN Analytics tool, detailing its novel methodology for temporal analysis, multi-view visualization (3D, Fitness View, 2D), and its integrated LLM-based assistant.
Camilo Chacón Sartori, Christian Blum and Gabriela Ochoa (2024). Large Language Models for the Automated Analysis of Optimization Algorithms. GECCO 2024.
Camilo Chacón Sartori, Christian Blum and Gabriela Ochoa (2024). An Extension of STNWeb Functionality: On the Use of Hierarchical Agglomerative Clustering as an Advanced Search Space Partitioning Strategy. GECCO 2024.
Camilo Chacón Sartori, Christian Blum and Gabriela Ochoa (2023). STNWeb: A new visualization tool for analyzing optimization algorithms. Software Impacts 2023.
Camilo Chacón Sartori, Christian Blum and Gabriela Ochoa (2023). Search Trajectory Networks Meet the Web: A Web Application for the Visual Comparison of Optimization Algorithms. ICSCA 2023.
Gabriela Ochoa, Katherine M. Malan and Christian Blum (2021). Search trajectory networks: A tool for analysing and visualising the behaviour of metaheuristics. Applied Soft Computing 2021.
Upload files with the format:
run_id,fitness,solution[,timestamp]
Generating STN, please wait...
Do you want to include the legend in the screenshot?
Choose where you want to place the legend.
| Algorithm |
Unique Nodes
|
Exclusive Nodes
|
Overall Best
|
Avg. Best Fitness
|
STD Best Fitness
|
Success Rate (%)
|
Avg. Path Length
|
Exploration Ratio
|
|---|
Algorithms to Compare
Comparing between and .
p-value
0.000
Configure and run a test to see the results here.
Opticat
Once you load your data, I can help you analyze it!
Your AI analysis partner
You have selected Hierarchical Clustering (HAC) for a large dataset (>5000 solutions). This process can be very slow and may cause the browser to become unresponsive.
For better performance, we strongly recommend using Mini-batch K-Means.
