The BMIPredict app makes a prediction from 2 individual variables (age and height) and 2 behavioral variables (the percentage of time spent in light activity during the day and the daily number of servings of vegetables consumed). These last 2 variables can be quantified using the WellBeNet application available on the PlayStore. Physical activity is determined precisely from the accelerometry data collected by the smartphone.
From the information entered in the application, the probability of belonging to the normal weight status, overweight or obese is presented in the form of histograms. The prediction comes from an algorithm produced by a neural network