Implementing supervised machine learning
The goal is to predict soccer match readiness score using supervised machine learning. Using data from the Oura ring (300 rows, so nearly a year's data), the model can predict the readiness score using three metrics: daily sleep score, daily activity score, and average resting heart rate.
Athletes tend to have a lower resting heart rate than non-athletes, so inevitably, their match readiness score will tend to be higher than the rest.
Ridge penalized coefficients with large positive or negative values. We need to choose the alpha value to fit or predict. We can select the alpha for which our model performs best.