How to quantify String Features while implementing Logistic Regression?

AI Discussions
1

While trying to build a Throat Cancer predictor with data containing 13 features as a very beginner project, I’m finding it confusing to utilize features like Gender which contain values ‘M’, ‘F’ etc. How do I inculcate this information while calculating the Loss function and how do I do the same for any other string valued features?

1 Like
2

You can convert the string features into one-hot binary values. For one simple example, for every text label, you’d create a feature, and its value would be either “True” or “False”.
The “True” values are 1, and the “False” values are 0.

3

Got it!
Thanks

1 Like