RMSE(Root Mean Square Error) is one of the most often indicators that measure the accuracy of artificial intelligence prediction models. To calculate RMSE, you need to first calculate the error which is the difference between the actual and predicted value. The mean of the squares of all errors is than calculated, which is then square rooted to find the RMSE. The equation below is the formula for RMSE, which is identical to the standard deviation formula.
- Inversely related to accuracy
RMSE value is a simple numeric indicater of a model's accuracy that is inversely related. The lower the RMSE value, the higher the accuracy.
- Scale dependant
RMSE is largely scale-dependant as its value can vary greatly depending on the scale of the prediction target. This property makes RMSE difficult to be used for comparing accuracy of multiple models with diferent scales.
CLICK AI's automated machine learning platform displays each model's RMSE to help users easily understand and find more accurate models.
[Generated models could be sorted according to RMSE, and more details are also available under the 'See details' tab.]