julsgasull/FIB-ADEI-LAB

Split the sample in work and test samples (consisting on a 80-20 split). Working data frame has to be used for model building purposes.

Outcome/Target

#67

· psclaudia opened

on Dec 7, 2020

At least two numerical variables have to be considered as explicative variables for initial steps in model building.

Outcome/Target

#66

· psclaudia opened

on Dec 7, 2020

Select the most significant factors according to feature selection as initial model factors. Put some reasonable limits to initial model complexity.

Outcome/Target

#65

· psclaudia opened

on Dec 7, 2020

You have to consider at least one interaction between a couple of factors and one interaction between factor and covariate.

Outcome/Target

#64

· psclaudia opened

on Dec 7, 2020

Diagnostics of the final model have to be undertaken. Lack of fit observations and influence data have to be selected and discussed (connections to multidimensional outliers in Multivariant Data Analysis is highly valuable).

Outcome/Target

#63

· psclaudia opened

on Dec 7, 2020

You have to predict Y (Binary Target) in the Working Data Frame vs the rest according to the best validated model that you can find and make a confusion matrix.

Outcome/Target

#62

· psclaudia opened

on Dec 7, 2020

Make a confusion matrix in the Testing Data Frame for Y (Binary Target) according to the best validated model found.

Outcome/Target

#61

· psclaudia opened

on Dec 7, 2020

Diagnostics of the final model have to be undertaken. Lack of fit observations and influence data have to be selected and discussed (connections to multidimensional outliers in Multivariant Data Analysis is highly valuable)

Multivariant Analysis conducted

#60

· psclaudia opened

on Dec 7, 2020

You have to consider at least one interaction between a couple of factors and one interaction between factor and covariate.

Multivariant Analysis conducted

#59

· psclaudia opened

on Dec 7, 2020

Select the most significant factors found in Multivariant Data Analysis as initial model factors. Put some reasonable limits to initial model complexity.

Multivariant Analysis conducted

#58

· psclaudia opened

on Dec 7, 2020

At least two numerical variables have to be considered as explicative variables for initial steps in model building, called covariates. Non-linear models have to be checked for consistency.

Multivariant Analysis conducted

#57

· psclaudia opened

on Dec 7, 2020

Economic vars

Explicative Variables for modeling purposes

#56

· psclaudia opened

on Dec 7, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Split the sample in work and test samples (consisting on a 80-20 split). Working data frame has to be used for model building purposes.

At least two numerical variables have to be considered as explicative variables for initial steps in model building.

Select the most significant factors according to feature selection as initial model factors. Put some reasonable limits to initial model complexity.

You have to consider at least one interaction between a couple of factors and one interaction between factor and covariate.

Diagnostics of the final model have to be undertaken. Lack of fit observations and influence data have to be selected and discussed (connections to multidimensional outliers in Multivariant Data Analysis is highly valuable).

You have to predict Y (Binary Target) in the Working Data Frame vs the rest according to the best validated model that you can find and make a confusion matrix.

Make a confusion matrix in the Testing Data Frame for Y (Binary Target) according to the best validated model found.

Diagnostics of the final model have to be undertaken. Lack of fit observations and influence data have to be selected and discussed (connections to multidimensional outliers in Multivariant Data Analysis is highly valuable)

You have to consider at least one interaction between a couple of factors and one interaction between factor and covariate.

Select the most significant factors found in Multivariant Data Analysis as initial model factors. Put some reasonable limits to initial model complexity.

At least two numerical variables have to be considered as explicative variables for initial steps in model building, called covariates. Non-linear models have to be checked for consistency.

Economic vars

Issues

Search results