obesity.levels: Obesity Levels

In fairml: Fair Models in Machine Learning

Obesity Levels

Description

Predict obesity levels based on eating habits and physical condition.

Usage

data(obesity.levels)

Format

The data contains 2111 observations and 17 variables. See the UCI Machine Learning Repository for details.

Note

The data set has been minimally pre-processed: the only change is that the only observation for which the CALC variable was equal to "Always" has been changed to "Frequently" to merge the two levels.

The obesity level NObeyesdad is the response variable (with 7 different levels) and Age and Gender are the sensitive attributes. The remaining variables are used as predictors.

The data contain the following variables:

• Gender;

• Age;

• Height;

• Weight;

• family_history_with_overweight;

• FAVC, frequent consumption of high caloric food as a factor with levels "no" and "yes";

• FCVC, frequency of consumption of vegetables as a numeric variable;

• NCP, number of main meals;

• CAEC, consumption of food between meals as a factor with levels "no", "Sometimes", "Frequently" and "Always";

• SMOKE, smoking status as a factor with levels "no" and "yes";

• CH2O, consumption of water daily as a numeric variable;

• SCC, calories consumption monitoring as a factor with level "no" and "yes";

• FAF, physical activity frequency as a numeric variable;

• TUE, time using technology devices as a numeric variable;

• CALC, consumption of alcohol as a dfactor with levels "no", "Sometimes", "Frequently" and "Always";

• MTRANS, transportation used as a factor with levels "Automobile", "Bike", "Motorbike", "Public_Transportation" and "Walking";

• NObeyesdad, the obesity level as a factor with levels "Insufficient_Weight", "Normal_Weight", "Overweight_Level_I", "Overweight_Level_II", "Obesity_Type_I", "Obesity_Type_II", "Obesity_Type_III".

References

UCI Machine Learning Repository.
https://archive-beta.ics.uci.edu/dataset/544

Examples

data(obesity.levels)

# short-hand variable names.
r = obesity.levels[, "NObeyesdad"]
s = obesity.levels[, c("Gender", "Age")]
p = obesity.levels[, setdiff(names(obesity.levels), c("NObeyesdad", "Gender", "Age"))]

## Not run: 
# the lambda = 0.1 is very helpful in making model estimation succeed.
m = fgrrm(response = r, sensitive = s, predictors = p, ,
      family = "multinomial", unfairness = 0.05, lambda = 0.1)
summary(m)

## End(Not run)

fairml documentation built on May 31, 2023, 6:02 p.m.