Name: Kiwser Ahmed Sajib

Section: 55-A
Id: 201-15-13876

The process of transforming unprocessed data into useful qualities for use in supervised learning is known as feature engineering. In order to extract useful characteristics that improve the results of machine learning, it depends on domain expertise. Brainstorming, feature selection, creation, testing, refining, and iteration are all parts of the process. Numerical transformations, category encoding (such as one-hot or target encoding), clustering, and aggregating values are typical strategies. The time-consuming activity of feature engineering includes data analysis, subject-matter expertise, and intuition. Its goal is to provide a set of useful and relevant features that improve a machine learning model's performance and accuracy.