Test predictions

IBM® Predictive Maintenance on Cloud analyzes historical data to discover patterns that are statistically significant as leading indicators of equipment failure. It uses these patterns as the basis for making predictions about the future.

The data that is used to make predictions is called the training data.

Sometimes patterns that are found in historical data are not likely to repeat. To avoid making predictions based on these patterns, Predictive Maintenance on Cloud sets aside data to test its ability to find patterns in unfamiliar data. This data is called the testing data.