tick: a Python library for statistical learning, with a particular emphasis on time-dependent modeling.
Team led by Prof. E. Bacry and Prof. S. Gaïffas in the Data Science Inititiative at Ecole polytechnique has just released tick, a machine learning library for Python 3. The focus is on statistical learning for time dependent systems. tick is used for many industrial applications including:
- A joint work with the French national health care (CNAMTS) to analyse the huge health database anonymously describing the medical care provided to more than 65 million beneficiaries. For this project, tick is used to detect weak signals in an unsupervised way such as unknown side effects for a given medicine.
- tick tools have been used in finance to model high-frequency order book data and analyse interactions between different event types and/or between different assets, leveraging the full time resolution available in the original data.
- Analyse social media information propagation. Thanks to a dataset collected during 2017 French election campaign on Twitter, tick is used to recover, for each topic, the network across which the information spread into the political spectrum.
The tick library is released with the support of Intel®. It uses the Intel® Math Kernel Library (MKL) optimized for Intel® Xeon Phi™ and Intel® Xeon™ processors. tick runs efficiently on everything from desktop computers to powerful high-performance servers.