TADkit: a toolkit for anomaly detection of time series data
How to create trustworthy timeseries data for the industry?
Industrial systems generate vast amounts of data, creating a need for industries to identify technologies that help experts annotate data with trustworthy information. TADkit is a software tools proposed by the foundation designed to improve data quality and identify biases in critical systems.
Why make it yours?
TADkit incorporates more than a dozen innovative methods into a single toolbox, such as deep neural network design, topological data analysis, and uncertainty quantification, to offer insights into the operation of complex systems across their lifecycle.
It allows users to configure, compare, and combine complex anomaly detection methods to enhance decision-making (« worth sending repair crew? »), reduce risks (early fault detection systems, monitoring systems) or improve operational efficiency (demand forecasting systems, environmental monitoring).
When to use it?
It could be used by Data Engineers or ML-Algorithm Engineers during the specification or the development of the ML Component or its models, and all along the data engineering life cycle.