The DataHour: An Overview of Feature Engineering for Data Science
Feature engineering is more than simply missing value imputation, handling outlier and categorical variables, and scaling numerical variables. It is an opportunity to allow a data scientist's creativity to shine and as #AndrewNg stated, “Applied machine learning is basically feature engineering.”
In this DataHour, Andrew had discussed advanced feature engineering techniques, best practices for ensuring these techniques work in a production environment, and finally, share a repository of more advanced feature engineering techniques.
Download the PDF: [ Ссылка ]
Part 2: [ Ссылка ]
Chapters
00:00 - 1:35: Introduction
1:36 - 2:28: Agenda
2:29 - 5:27: Why Feature Engineering
5:28 - 8:32:What is Feature Engineering
8:33 - 19:30: Best practices
19:31 - 30:13: Examples
30:14 - 31:02: One to Many Relationships
31:03 - 32:23: Date Features
32:33 - 36:50: Numeric Features
36:51 - 37:44: Categorical Features
37:45 - 40:47: Group Features, Lag Features
40:48 - 46:15: This is the Fun part
46:16 - 59:50:Q/A
🔗 More action pack session here: [ Ссылка ]
Stay on top of your industry by interacting with us on our social channels:
Follow us on Instagram: [ Ссылка ]
Like us on Facebook: [ Ссылка ]
Follow us on Twitter: [ Ссылка ]
Follow us on LinkedIn: [ Ссылка ]
