Session Ready
Exercise

Fitting logistic models

Many business problems require the prediction of a binary response variable. Your future employer may need to detect spam e-mails, credit card frauds, or rare diseases.

The logistic regression model is the go-to method for binary classification problems.

In this exercise, you will use Parkinson's dataset from the UCI repository. This dataset is composed of a range of biomedical voice measurements from people with and without Parkinson's disease.

You will use the following variables from the dataset:

  • status - 1 - if a person has Parkinson's disease, 0 - otherwise,
  • NHR - a measure of the ratio of noise to tonal components in the voice,
  • DFA - a signal fractal scaling exponent.

The dataset is available as parkinsons.

Instructions 1/2
undefined XP
  • 1
  • 2
  • Draw a scatterplot of status against NHR from the parkinsons dataset.
  • Draw a scatterplot of status against DFA from the parkinsons dataset.