Easy Fit / Easy Predict

Easy Fit allows you to test and compare different predictive models for the same dataset. Depending on the type of the dependent and explanatory variables (quantitative or qualitative), various models are proposed. The Easy Predict function can be then used to make predictions using the previously fitted models.

Options for the Easy Fit /  Easy Predict function in XLSTAT

Available regression models

Easy Fit offers the following regression models:

• If the X explanatory variables are only quantitative:

  • Linear regression

  • Regression random forests

  • K nearest neighbors

  • Support vector machines

  • LASSO regression

  • Extreme gradient boosting (XGBOOST)

• If the X explanatory variables are only qualitative:

  • Analysis of variance: ANOVA

  • Regression random forests

  • K nearest neighbors

  • Support vector machines

  • LASSO regression

  • Extreme gradient boosting (XGBOOST)

• If some X explanatory variables are quantitative and other qualitative:

  • Analysis of covariance analysis: ANCOVA

  • Regression random forests

  • K nearest neighbors

  • Support vector machines

  • LASSO regression

  • Extreme gradient boosting (XGBOOST)

Available classification models

Easy Fit offers the following classification models (you can click on the different methods to access the associated help document):

For any type of explanatory variables X, whether only quantitative, only qualitative or both quantitative and qualitative:

• Logistic regression

• Classification random forests

• K nearest neighbors

• Support vector machines

• Discriminant analysis

• Extreme gradient boosting (XGBOOST)

Results for the Easy Fit /  Easy Predict function in XLSTAT

Descriptive statistics: The tables of descriptive statistics show basic statistics for all the selected variables such as the number of observations, missing values, the number of non-missing values, the mean and the standard deviation (unbiased) are displayed for the quantitative variables. For qualitative variables the names of the various categories are displayed together with their respective frequencies.

Summary table: The summary table contains the following indicators: MAE, MSE, R², adjusted R², AIC and SBC. Therefore for a classification, the indicators are: Accuracy, Precision, Recall, Correct classification number, Misclassification number and F-score (see Model Performance Indicators).

Summary results by method: For each method the most important results of the method are displayed. For a better understanding of the various results, you can access the help section of the related methods.