How to Lower the Root Mean Square Value

The root mean square error (often abbreviated RMSE) is a measure of the amount of error in a statistical model. It is calculated by first finding the distance between each predicted value of the dependent variable from the model and the actual value from the data, then squaring these distances, then summing the squares and finally taking the square root of the sum.

Instructions

    • 1

      Add more variables to the model. One method of reducing the root mean square error is to make a more complex model. While more complex models always fit the data better, the value of this better fit must be weighed against the added complexity.

    • 2

      Add transformations of the variables to the model. For example, if you are modelling weight as a function of height in adult humans, it will reduce root mean square error if you add height squared to the model. However, the same cautions apply regarding complexity of the model.

    • 3

      Get more accurate data. If you can measure your variables more precisely, then this will reduce root mean square error. Part of the error that is measured by RMSE is due to inaccuracies of the model; no model is perfect. But any inaccuracy in measurement will also add to the error.

Learnify Hub © www.0685.com All Rights Reserved