Implementing logistic regression

For this recipe, we will implement logistic regression to predict the probability of low birth weight in our sample population.

Getting ready

Logistic regression is a way to turn linear regression into a binary classification. This is accomplished by transforming the linear output into a sigmoid function that scales the output between zero and one. The target is a zero or one, which indicates whether a data point is in one class or another. Since we are predicting a number between zero and one, the prediction is classified into class value 1 if the prediction is above a specified cutoff value, and class 0 otherwise. For the purpose of this example, we will specify that cutoff to be 0.5, which will make the classification as simple as rounding the output.

The data we will use for this example will be the low birth weight data obtained from the author's GitHub repository (https://github.com/nfmcclure/tensorflow_cookbook/raw/master/01_Introduction/07_Working_with_Data_Sources...