Celia Lozano GrijalbaTech Lead Data Scientist at Bosonit, shares with us the Sktime projectThe project is oriented towards solving time series data science problems in Python.
Solving problems of Data Science with time series in Python is a challenge. Why? Existing tools are not well suited to time series tasks, nor are they easily integrated. The models in the scikit-learn package assume that the data is structured in a tabular format. It also assumes that each column is id, assumptions that do not apply to time series data. Packages containing time series learning modules, such as statistical models, do not integrate well with each other.
In addition, many essential time series operations, such as splitting data into test sets and trains over time, are not available in existing Python packages. To address these challenges, the following was created Sktime.
Sktime is a community-driven project funded by the UK Economic and Social Research Council, the Consumer Data Research Centre and the Alan Turing Institute (where Celia had the opportunity to work during her time in Manchester).
This project extends the scikit-learn API to time series tasks. It provides the algorithms and transformation tools needed to efficiently solve time series classification, forecasting and regression tasks.
If you want to know more about Sktime. You can find out more about Sktime at the following link:
Author: Celia Lozano GrijalbaTech Lead Data Scientist at Bosonit.