Implementations, mostly pedagogical, of various algorithms and data-structures.
The domains of interest range from classic algorithms and data-structures to machine-learning and AI through quantitative finance and distributed systems.
The primary intention of this library is to develop, in Python and using only the standard library and common external libraries (numpy, pandas, for example), simple implementations of algorithms in common use. Often, but not always, there is a better performing implementation available in open-source or in a higher performance language. My intention here is not to replace them, but to simplify them. Together, the documentation and the code should yield something of a clear pedagogical narrative.
The documentation is available here : https://inegm.github.io/toolbox/
This library will be made available on PyPI once I've settled on a decent number of modules for a first package.
- Artificial Neural Networks
- Reinforcement Learning (Q-learning)
- Ant-colony optimization
- Evolutionary algorithms for optimization
- Financial market-risk metrics (VaR, ES, etc)
- Various auction mechanisms
- MCMC methods
- GARCH processes
- Clock synchronization (from the SNTP)
- Best master clock (from the PTP)