Add basic Machine Learning

- Clust Algorithms
- Predictions when network is about to fail

The issue mentions adding clustering algorithms and failure prediction, but does not specify where to start, what data to use, or how to integrate with existing network analysis scripts. Look at the repository's existing scripts to understand the data format and network metrics collected. Determine what constitutes a 'network failure' and how to model it. This is a large, open-ended feature requiring significant design and implementation.

索引モデルが issue の本文から書いたものです。