This is the R (Microsoft R Server) code for Predictive Maintenance template using SQL Server R Services.
The template solves the following problems:
- Predict the Remaining Useful Life (RUL) of an asset, or Time to Failure (TTF). This is formulated as a regression problem.
- Predict if an asset will fail within certain time frame (e.g. days). This is formulated as a binary classification problem.
- Predict if an asset will fail in different time windows. This is formulated as a Multi-class classification problem.
It consists of the following files:
| File | Description |
|---|---|
| 01-data-preparation.R | Load data to SQL tables, data labeling, feature engineering, normalization |
| 02a-regression-modeling | Train and evaluate multiple regression models |
| 02b-binary-classification-modeling | Train and evaluate multiple binary classfication models /td> |
| 02b-binary-classification-modeling | Train and evaluate multiple multiclass classfication models /td> |
A detailed description of the template, implemented in Azure Machine Learning Studio can be found here.