#featureengineeering #machinelearning #datascience
Feature engineering is a process of taking unrefined raw data and transforming it in meaningful way such that model can understand the data better
Logic on how to transform a model can come from Domain understanding of the business or Data Analysis or EDA cycle
3 key benefits of feature engineering include
- Feature engineering can keep your model simple. As simpler algorithm can fine better pattern in data
- Provide better explain-ability
- Remove human bias by modelling some critical data proxies as unbiased features
