Home

Introduction

pyclustring is a python data mining library which is focused on bio-inspired algorithms, such as algorithms based on oscillatory neural networks, and on classical algorithms that are widely used in many applications. pyclustring provides opportunity to examine oscillatory networks and neural networks abstracting from data mining. pyclustring can be used in python scripts as well as in others languages using dynamic link library of pyclustering for windows or linux.

Implemented essences

pyclustering implements data mining algorithms, models of oscillatory networks and neural networks. Each implemented essence contains example of using and demonstration of basic features.

Clustering algorithms:

• BIRCH
• CURE
• DBSCAN
• Hierarchical
• HSyncNet (bio-inspired)
• K-Means
• OPTICS
• ROCK
• SyncNet (bio-inspired)
• SyncSom (bio-inspired)
• X-Means

Oscillatory and neural networks models:

• HHN (Oscillatory network based on Hodgkin-Huxley model)
• Hysteresis Oscillatory Network
• LEGION (Local Excitatory Global Inhibitory Oscillatory Network)
• PCNN (Pulse-Coupled Neural Network)
• SOM (Self-Organized Feature Map)
• Sync (Oscillatory network based on Kuramoto model)

Graph coloring algorithms:

• DSatur
• Hysteresis (bio-inspired)
• Sync (bio-inspired)

** Containers **

• KD Tree
• CF Tree