This algoritm was implemented in Jupyter notebook in Python 3.7 for prediction maintenance on the chimical factory.
We have data from last year, we generated new data for more years and algoritm was learning using Random Forest Classification to indicate and predictive maintenance.
Entry: data by last day ( 24 hours ) Exit: Yes/No failure after 40-43 hours
On the test data the result have accuracy 99.9 %
prediction maintenance.ipython - code with algorithm Shufling_maintenance.xlsx - generated data for learning