Model fitting
Model fitting refers to the process that takes three steps; the first step is to have a function that takes in a set of parameters and returns a data set that is predicted. The second step is to have an error function that provides a number representing the difference between the provided data and the model’s prediction. The last step is to identify the parameters that minimize the difference noted. A well-fitting model regression model results in predicted values. It is, therefore, essential to ensure that the data model fits well with a given set of data (Miles, 2014). This can be achieved by having effective criteria in place that will help in monitoring and checking if there is a good fitting.
Some of the strategies that can be used in the evaluation process include R squared and adjusted E-squared. R-square has a useful property that its scale is intuitive, meaning that it ranges from zero to one. The zero indicates that the proposed model does not improve prediction over the mean model. On the other hand, one depicts that the prediction is perfect. This is one of the best strategies that can be employed to help in evaluating how well a model fits data. The only problem with R-square is that it can only increase as predictors are added to the regression model, and to resolve this problem, adjusted E –square provides the freedom needed to deal with this issue (Ricci, 2010). Adjusted R-square helps in decreasing as the predictors are added. Through this method, it is easy for one to identify if the model can fit well with data.