Share
What is Feature Engineering? Explain with Example?
Question
TheDataMonk
Lost your password? Please enter your email address. You will receive a link and will create a new password via email.
Answers ( 2 )
Feature engineering is the process of using domain knowledge of the data to
create features for machine learning algorithm to work
-Adding more columns (or) removing columns from the existing column
-Outlier Detection
-Normalization etc
Feature engineering is the ? of your problem’s solution. If you have done feature engineering in best way then you will reach to heart.
Coming into the formal definition, Feature engineering is part where you use your domain knowledge in problem to identify the relevant features and create them from the existing feature.
For example you are doing forecasting sales problem. You are having date variable then you can create the day of year, day of month, month, day of week etc time based features. Apart from this you can also create some lag features. Let’s say you have to predict for next day then you can create lag features using what was last day price. (I am taking when you are using machine learning model for this problem, there are some stats models which can help without these lag variables in forecasting problem)