dsflow is a repository of typical workflows for Data Science in Python. Its purpose is to provide Data Scientists with a way to quickly bootstrap solutions to typical problems such as 'likelihood to x' which can be solved with binary classification. This repo will include clustering soon, please suggest more under Issues!.
- Template workflows and library for typical machine learning problems for Data Scientists, for example:
- 'likelihood to VIP'
- 'likelihood to churn'
- Template takes care of:
- Data cleanup, feature selection, model training, evaluation, cross-validation, etc.
- Training multiple models and provides comparative visualizations.
- Jump start a new machine learning project!
- Easily share with others.
- Learn Python, machine learning, and Data Science by example.
- Interface is an interactive Jupyter Notebook.
- Built using popular Python libraries: SciKit Learn, Pandas, NumPy, MatPlotLib.